Optimized operation method, system and device based on multiple distributed power supplies and main and distribution networks
By building the optimal trend model of distribution network and distributed power supply, using Ward equivalent theory and saddle point dynamic method, the distributed optimal trend optimization of distribution network and distributed power supply is achieved, and the problem of inapplicability of traditional regulation strategies is solved. , verified the effectiveness and accuracy of the method.
Patent Information
- Application Number
- CN202411123832.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-16
- Publication Date
- 2024-11-26
AI Technical Summary
How to achieve decentralized collaborative optimization of distribution networks and distributed power supplies, and solve the problem that traditional centralized unified regulation strategies are no longer applicable.
By constructing the optimal current model of distribution network and distributed power supply, using Ward equivalent theory to decouple the coupling network equation, construct a quadratic planning model, and using saddle point dynamic method for optimization and solution, realizing the distribution network and distributed power supply information exchange and optimal trend dispersed solution.
Effectively solve the problem that the existing regulatory strategies cannot adapt to the active distribution system containing multiple distributed power supplies, construct a linear AC current equation suitable for distribution networks and distributed power supplies, and realize the dispersion between distribution networks and distributed power supplies. The optimal trend optimization verifies the effectiveness and accuracy of the method.
Smart Images

Figure CN119029855A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power distribution system control, and in particular to a method, system and device for optimizing operation based on multiple distributed power sources and a main distribution network. Background Art
[0002] The access of renewable energy sources such as photovoltaic and wind power provides more available resources for the distribution network side. However, distributed power sources such as photovoltaic and wind power are seriously affected by meteorological factors and have the characteristics of volatility and randomness, which greatly restricts the scale of distributed power access and energy utilization. At present, energy storage devices are added to power generation, and the charging and discharging characteristics of energy storage devices are used to reduce the fluctuation of electric energy.
[0003] In the current process of electricity sales reform, with a large number of distributed renewable energy sources connected to the end of the distribution network in the form of distributed power sources, distributed power sources include a mixture of multiple power sources such as photovoltaics, wind power and energy storage devices, and use internal control to achieve stable power output and meet internal load requirements with the goal of cost and renewable energy consumption. Therefore, they can be used as separate independent distributed power sources, and can be operated independently as a subject that does not belong to the distribution system. Therefore, the output and flow information of its internal components should not be obtained by the distribution network operator. Similarly, the information within the distribution network should not be transmitted to the distributed power source operator. This makes the traditional centralized unified control strategy no longer applicable to active distribution systems containing multiple distributed power sources. Therefore, how to achieve decentralized collaborative optimization of distribution networks and distributed power sources has become an important research issue. Summary of the invention
[0004] The problem solved by the present invention is: how to achieve decentralized collaborative optimization of distribution networks and distributed power sources.
[0005] To solve the above problems, an embodiment of the present invention provides an optimization operation method based on multiple distributed power sources and a main distribution network, characterized in that the optimization operation method includes: obtaining the power flow conditions of the distribution network in the target area, and constructing an optimal power flow model of the distribution network according to the power flow conditions, which is recorded as the distribution network optimal model; obtaining the grid-connected power flow conditions of the distributed power sources in the target area, and constructing an optimal power flow model of the distributed power sources according to the grid-connected power flow conditions, which is recorded as the power supply optimal model; decoupling the original coupling network equations of the distribution network and the distributed power sources through the Ward equivalence theory; constructing a quadratic programming model for solving the decentralized optimal power flow according to the decoupling results; and optimizing and solving the quadratic programming model through the saddle point dynamic method.
[0006] Furthermore, the power flow conditions of the distribution network are obtained in the target area, and the optimal power flow model of the distribution network is constructed according to the power flow conditions, which is recorded as the optimal distribution network model, specifically including:
[0007] Taking network loss as the optimization target, the optimal distribution network model is constructed under the conditions of satisfying linearized AC power flow constraints, node voltage constraints and generator output constraints.
[0008] Furthermore, the grid-connected power flow of distributed power sources is obtained in the target area, and the optimal power flow model of distributed power sources is constructed according to the grid-connected power flow, which is recorded as the power source optimal model, specifically including:
[0009] The optimization objectives are network loss costs, conventional generator operating costs, power abandonment costs and energy storage charging and discharging loss costs, and an optimal power supply model is constructed under the conditions of satisfying linearized AC power flow constraints, node voltage constraints, generator output constraints, renewable energy generation constraints and energy storage system power constraints.
[0010] Furthermore, the original coupling network equations of the distribution network and distributed power sources are decoupled through the Ward equivalence theory, specifically including: using the Ward equivalence theory, the network and power information inside the distributed power source is equivalent to an equivalent injected power on the boundary node; after completing the Ward equivalence, the distributed power source transmits the equivalent power as boundary information to the distribution network; and obtains the boundary coupling constraints when the distributed power source is equivalent.
[0011] Furthermore, the original coupling network equations of the distribution network and distributed power sources are decoupled through the Ward equivalence theory, specifically including: using the Ward equivalence theory, firstly equate the network and power information of the distributed power source to a node inside the distribution network to inject power, and then equate the network and power information of the distribution network to an equivalent injected power on the boundary node; when the distribution network and the distributed power source complete the Ward equivalence, the equivalent power is transmitted to the distributed power source as the boundary information; and the boundary coupling constraints when the distribution network is equivalent are obtained.
[0012] Furthermore, a quadratic programming model for solving the decentralized optimal power flow is constructed based on the decoupling results, specifically including: constructing a quadratic programming model for solving the decentralized optimal power flow based on the optimal distribution network model, the optimal power supply model and the boundary coupling constraints during the equivalence.
[0013] Furthermore, the quadratic programming model is optimized and solved through the saddle point dynamic method, specifically including: constructing an augmented Lagrangian function for the quadratic programming model for solving the decentralized optimal power flow, and according to the principle of the saddle point dynamic method, from the perspective of dynamic control stability, converting the KKT condition of an original convex optimization problem into an asymptotic stability process of a dynamic feedback control system, and obtaining the optimal solution after solving the KKT condition.
[0014] Furthermore, the present application also provides a multi-distributed power supply and main distribution network optimization operation system, the multi-distributed power supply and main distribution network optimization operation method recorded in the above embodiment is applied to the operation system, and the operation system includes: a distribution network model construction module: taking network loss as the optimization target, and constructing the optimal distribution network model under the conditions of satisfying the linearized AC power flow constraints, node voltage constraints and generator output constraints; a distributed power supply model construction module: taking network loss costs, conventional generator operating costs, power abandonment costs and energy storage charging and discharging loss costs as the optimization targets, and constructing the optimal power supply model under the conditions of satisfying the linearized AC power flow constraints, node voltage constraints, generator output constraints, renewable energy generation constraints and energy storage system power constraints; a planning model construction module: using Ward equivalence theory to optimize the distributed The network and power information inside the power source is equivalent to an equivalent injected power on the boundary node; after the distributed power source completes the Ward equivalence, the equivalent power is transmitted to the distribution network as the boundary information; and the boundary coupling constraints when the distributed power source is equivalent are obtained; using the Ward equivalence theory, the network and power information of the distributed power source is first equivalent to a node injection power inside the distribution network, and then the network and power information of the distribution network is equivalent to an equivalent injected power on the boundary node; when the distribution network and the distributed power source complete the Ward equivalence, the equivalent power is transmitted to the distributed power source as the boundary information; and the boundary coupling constraints when the distribution network is equivalent are obtained; a quadratic programming model for solving the decentralized optimal power flow is constructed based on the optimal model of the distribution network, the optimal model of the power source and the boundary coupling constraints when performing the equivalence. Optimization module: construct an augmented Lagrangian function for solving the quadratic programming model of decentralized optimal power flow. According to the principle of the saddle point dynamic method, from the perspective of dynamic control stability, the KKT condition of an original convex optimization problem is converted into an asymptotic stability process of a dynamic feedback control system. After solving the KKT condition, the optimal solution is obtained. The operating system has all the technical features of the above-mentioned optimization operation method, which will not be described one by one here.
[0015] The beneficial effects of the present invention are:
[0016] 1. The present invention solves the problem that the existing control strategy cannot adapt to the active distribution system containing multiple distributed power sources by constructing models separately and using Ward equivalence theory to realize information exchange between the two and form a planning model.
[0017] 2. The present invention constructs a Ward equivalent model suitable for the linearized AC power flow equation of the distribution network and distributed power sources, decouples the original coupling network equations of the distribution network and distributed power sources, and further constructs a quadratic programming model for solving the decentralized optimal power flow. The saddle point dynamic method is used to transform the optimization process of the above quadratic programming model into an asymptotically stable process of a dynamic feedback control system. The proposed model was simulated and tested in two actual distribution network systems containing multiple distributed power sources, verifying the effectiveness and accuracy of the proposed decentralized solution method. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 This is a flow chart of the method for optimizing the operation of the main distribution network for this application;
[0019] Figure 2 The feedback control system diagram of the saddle point dynamic method;
[0020] Figure 3 The topological structure diagram of the test system for this application;
[0021] Figure 4 This is a trajectory curve diagram of the voltage amplitude of this application;
[0022] Figure 5 is an output power curve diagram of each component in the distributed power supply 1;
[0023] Figure 6 is an output power curve diagram of each component in the distributed power source 2;
[0024] Figure 7 This is the interactive power trajectory curve between the distribution network and the microgrid in the test system
[0025] Figure 8 This is the system schematic diagram of the main distribution network optimization operation system for this application.
[0026] Description of reference numerals:
[0027] 100 - operation system; 110 - distribution network model building module; 120 - distributed power supply model building module; 130 - planning model building module; 140 - optimization module. DETAILED DESCRIPTION
[0028] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.
[0029] [First embodiment]
[0030] See also Figure 1In a specific embodiment, the present invention provides an optimization operation method based on multiple distributed power sources and a main distribution network, characterized in that the optimization operation method includes:
[0031] S100, obtaining the power flow condition of the distribution network in the target area, and constructing an optimal power flow model of the distribution network according to the power flow condition, which is recorded as the distribution network optimal model;
[0032] S200, obtaining the grid-connected power flow of the distributed power source in the target area, and constructing an optimal power flow model of the distributed power source according to the grid-connected power flow, which is recorded as the power source optimal model;
[0033] S300, decouple the original coupling network equations of the distribution network and distributed generation through Ward equivalence theory;
[0034] S400, constructing a quadratic programming model for solving a decentralized optimal power flow according to the decoupling result;
[0035] S500, optimizing and solving the quadratic programming model through the saddle point dynamic method.
[0036] In step S100, the target area is usually a city, but can also be a village or an administrative planning area. The power grid flow situation is obtained through the data management library, and the distribution network optimal model is calculated by the following formula:
[0037] f DS =ρ loss ·F D-loss +ρ g ·F D-g (1);
[0038]
[0039] In the above formula, F DS represents the optimization objective function of the distribution network operator, F D-loss represents the network loss cost of the distribution network, F D-g represents the power generation cost of the distribution network, ρ loss and ρ g represents the corresponding objective function weight coefficient, c loss represents the unit network loss cost coefficient, a g The quadratic coefficient of the generator operating cost, b g The linear coefficient representing the generator operating cost, represents the set of lines in the distribution network, represents the set of generators in the distribution network, y ij represents the admittance between nodes i and j, Re{·} represents the real part of the value in brackets, e i and fi They represent the real and imaginary parts of the voltage vector at node i, respectively, P i g Indicates the active output value of the generator.
[0040] Furthermore, the linearized AC power flow constraint is as follows:
[0041]
[0042] Among them, N s represents the total number of nodes in the distribution network, represents the set of distribution network nodes, V i and θ i represents the voltage amplitude and phase angle of node i at time t, P i and Q i represents the active power and reactive power injected by node i, with the power generation direction as the positive direction. and represents the impedance matrix element value only related to the distribution network node after removing the balancing node, P i g represents the active power of the generator at node i, represents the reactive power of the generator at node i, P i l and The active and reactive power of the load at node i respectively.
[0043] The voltage of each node in the distribution network should be kept within the limited range. Assuming that the voltage amplitude of the balance node is 1p.u and the voltage phase angle is 0, the upper and lower limit constraints of the voltage of each node meet the following constraints:
[0044] V min ≤V i ≤V max ,
[0045] V0=1,θ0=0 (9).
[0046] Among them, V min and V max They represent the upper and lower limits of the voltage amplitude at node i, respectively, and V0 and θ0 represent the voltage amplitude and phase angle of the balanced node, respectively.
[0047] The upper and lower limits of the active and reactive output of the generator are both fixed values, and the specific constraints are as follows:
[0048]
[0049] in, Indicates the maximum active output of the generator. Indicates the maximum reactive power output of the generator.
[0050] In step S200, the process of constructing the optimal power model is: taking network loss costs, conventional generator operating costs, power abandonment costs and energy storage charging and discharging loss costs as optimization targets, and constructing the optimal power model under the conditions of satisfying linearized AC power flow constraints, node voltage constraints, generator output constraints, renewable energy generation constraints and energy storage system power constraints, as shown in the following formula:
[0051]
[0052] Among them, F MG represents the optimization objective function of the distributed generation operator, represents the network loss cost of distributed generation, represents the penalty cost of curtailment of renewable energy, represents the operating cost of a conventional generator, represents the charging and discharging loss cost of the energy storage battery system, ρ loss , rer , g and ρ ess represents the corresponding objective function weight coefficient, represents the set of lines in the distributed power generation, represents the set of conventional generators in the distributed generation, Represents a collection of renewable energy generation equipment in distributed power sources. represents the set of energy storage battery systems in distributed power sources, P i c represents the amount of abandoned electricity generated by renewable energy, a r represents the quadratic coefficient of the power curtailment penalty cost, b r The linear coefficient of the penalty cost for power curtailment, P i chi and P i disi Respectively represent the charging power and discharging power of the energy storage battery system, a ch and a dis They represent the charging loss coefficient and the discharging loss coefficient respectively.
[0053] Furthermore, the linearized AC power flow constraint is as follows:
[0054]
[0055] Among them, N m represents the total number of nodes in the distributed generation, represents the collection of distributed generation nodes, and represents the impedance matrix element value related only to the distributed generation node, P i der Represents the maximum output active power of renewable energy generation, represents the reactive power of the renewable energy inverter at node i, represents the reactive power of the conventional generator at node i.
[0056] The voltage of each node in the distributed power supply should be kept within the limited range, and the upper and lower limits of the voltage of each node should meet the following constraints:
[0057] V min ≤V i ≤V max ,
[0058] Among them, V min and V max They represent the upper and lower limits of the voltage amplitude at node i, respectively, and V0 and θ0 represent the voltage amplitude and phase angle of the balanced node, respectively.
[0059] The upper and lower limits of the active and reactive output of the generators in the distributed power generation are both fixed values, and the specific constraints are as follows:
[0060]
[0061] in, Indicates the maximum active output of the generator. Indicates the maximum reactive power output of the generator.
[0062] Renewable energy generation constraints: Renewable energy mainly includes wind power and photovoltaic power. Its maximum active output is affected by the external environment (such as light, wind speed, etc.) and exhibits time-varying properties, while its reactive output can be adjusted by the inverter. Therefore, the corresponding constraints are as follows:
[0063]
[0064] in, Indicates the power factor angle of the inverter.
[0065] Energy storage system power constraint: For the energy storage capacity at the end of the previous period, it can be regarded as a known input parameter W i,0 , therefore, the power constraint of the energy storage system in the current period can be obtained:
[0066]
[0067] Among them, P i ch and P i disRespectively represent the maximum charging power and maximum discharging power of the energy storage battery under rated conditions, W i,max and W i,min Respectively represent the upper and lower limits of the energy storage battery, η c and η d Represent the charging efficiency and discharging efficiency respectively.
[0068] In step S300, since multiple distributed power sources are connected to the distribution network, and the distributed power sources and the distribution network are independent of each other and do not affect each other, the grid structure and node power information are isolated from each other. At the same time, in order to protect the privacy data such as internal power information, voltage information and grid information of different entities, the network equations between the distribution network and the distributed power sources are decoupled by the Ward equivalence method, so that the distribution network and the distributed power sources only need to exchange boundary information to achieve the decentralized solution of the optimal power flow.
[0069] The process of decoupling the original coupled network equation of the distributed power source is as follows: using Ward's equivalence theory, the network and power information inside the distributed power source is equivalent to an equivalent injected power on the boundary node; the specific expression of the equivalent power of the distributed power source at the boundary node is as follows:
[0070]
[0071] Among them, B M represents the boundary node set on the distributed generation side, P M and Q M They represent the active power and reactive power vectors injected into the nodes of the distributed generation, and Represents the node impedance vector between the distributed generation boundary node and the distributed generation internal node.
[0072] After completing the Ward equivalent, the distributed generation transmits the equivalent power as boundary information to the distribution network; at this time, the boundary node replicated by the distribution network injects power and Required and actual equivalent power and The following consistency constraints are formed:
[0073]
[0074] Similarly, the voltage amplitude and phase angle at the boundary node must also satisfy the consistency constraint:
[0075] e s =e′ s ,
[0076] f s =f′s ,
[0077] Among them, the process of decoupling the original coupled network equation of the distribution network is as follows: using the Ward equivalence theory, the network and power information of the distributed power source is firstly equivalent to a node injection power inside the distribution network, and then the network and power information of the distribution network is equivalent to an equivalent injection power on the boundary node; the specific expression of the equivalent power of the distribution network and other distributed power sources at the boundary node is as follows:
[0078]
[0079] In the formula, B D represents the boundary node set on the distribution side, P D and Q D They represent the active power and reactive power vectors injected into the nodes in the distribution network, and It represents the node impedance vector between the distribution network boundary nodes and the internal nodes of the distribution network.
[0080] When the distribution network and the distributed power source complete the Ward equivalence, the equivalent power is transmitted to the distributed power source as the boundary information; at this time, the boundary node replicated by the distributed power source injects power and Required and actual equivalent power and The following consistency constraints are formed:
[0081]
[0082] Similarly, the voltage amplitude and phase angle at the boundary node must also satisfy the consistency constraint:
[0083] e r =e′ r ,
[0084] f r =f′ r ,
[0085] The process of constructing a quadratic programming model for solving decentralized optimal power flow is as follows: a quadratic programming model for solving decentralized optimal power flow is constructed based on the optimal model of the distribution network, the optimal model of the power source and the boundary coupling constraints when performing equivalence.
[0086] The compact form of the distribution network in the quadratic programming model of decentralized optimal power flow is as follows:
[0087]
[0088] In the formula, xd represents the optimization variables in the distribution network, mainly including the voltage amplitude and phase angle of each node and the active and reactive power of the generator. Formula (40) represents the optimization variables about x d The quadratic objective function, formula (41) represents the linearized AC power flow equation related only to the distribution network nodes, formula (42) represents the various inequality constraints in the distribution network, mainly the upper and lower limit constraints of the generator and the upper and lower limit constraints of the node voltage, and formula (43) represents the boundary coupling constraints between the distribution network and the distributed generation; in the distribution network problem It represents the boundary node voltage information and equivalent injected power information transmitted from the distributed power source and can be regarded as a constant value.
[0089] The compact form of distributed generation in the quadratic programming model of decentralized optimal power flow is as follows:
[0090]
[0091] In the formula, x m represents the optimization variables in the mth distributed power source, which mainly include the voltage amplitude and phase angle of each node, the active and reactive power of the generator, the abandoned power of renewable energy generation and the reactive power of the inverter, and the charging power and discharging power of the energy storage system. Formula (44) represents the optimization variables about x m The quadratic objective function, formula (45) represents the linearized AC power flow equation related only to the internal nodes of the distributed power source, formula (46) represents the inequality constraints in the distributed power source, and formula (47) represents the boundary coupling constraint between the mth distributed power source and the distribution network; in the distributed power source problem It represents the boundary node voltage information and equivalent injected power information transmitted from the distribution network and can be regarded as a constant value.
[0092] In the above-mentioned step S4, the process of optimization and solution is as follows: an augmented Lagrangian function is constructed for the quadratic programming model for solving the decentralized optimal power flow, and according to the principle of the saddle point dynamic method, from the perspective of dynamic control stability, the KKT condition of an original convex optimization problem is converted into an asymptotic stability process of a dynamic feedback control system, and after solving the KKT condition, the optimal solution is obtained.
[0093] Furthermore, for the distribution network problem and distributed power supply problem, the augmented Lagrangian function is constructed as follows:
[0094]
[0095] In the formula, m represents the total number of distributed generation, λ d and μ d The Lagrange multipliers corresponding to the equality constraints (41) and inequality constraints (42) of the distribution network subproblem, λ m and μ mThe Lagrange multipliers corresponding to the equality constraints (45) and inequality constraints (46) of the mth distributed power problem, λ co,m Represents the Lagrange multiplier corresponding to the boundary coupling constraints of the distribution network and distributed generation.
[0096] The differential equations for the dynamic evolution of the primal and dual variables of the distribution network problem are as follows:
[0097]
[0098] The differential equations for the dynamic evolution of the primal and dual variables of the distributed power problem are as follows:
[0099]
[0100] The differential equation for the dynamic evolution of the dual variables corresponding to the boundary coupling constraint is as follows:
[0101]
[0102] It can be found that when the distribution network and distributed generation update their internal variables respectively, they no longer need to know the global information, but only need to exchange the voltage information and equivalent power information at the boundary nodes.
[0103] The KKT condition of an original convex optimization problem is converted into an asymptotically stable process of a dynamic feedback control system: The dynamic process of the saddle point dynamics method can be described as a dynamic feedback control system, see Figure 2 . Among them, the dotted box 1 reflects the gradient generation process of the original variable x and the proportional-integral dynamic process to eliminate the steady-state deviation; the dotted boxes 2 and 3 represent the proportional-integral dynamic process of the Lagrange multipliers λ and μ respectively. The above proportional-integral links are coupled according to the above differential equations to form a feedback control system. At this point, the KKT condition of an original convex optimization problem can be converted into an asymptotically stable process of a dynamic feedback control system.
[0104] The optimization configuration method proposed in this application is verified below.
[0105] A 35-node actual distribution system with two independent distributed generation sources is selected to verify the effectiveness of the proposed method. Figure 3 As shown in the figure, there are 35 distribution nodes and 37 feeders. The dotted lines represent the tie lines with switches, which are open in normal operation. The voltage level of the distribution system is 10.5kV, the reference power is 100MW, the distributed generation 1 (MG1) is connected to the node 12, and the distributed generation 2 (MG2) is connected to the node 33. The internal nodes of the distributed generation 1 are numbered from 36 to 43, and the internal nodes of the distributed generation 2 are numbered from 44 to 56.
[0106] Distributed power source 1 includes a reactive power compensator, a photovoltaic power generation group, a battery storage system and a group of loads; distributed power source 2 includes a photovoltaic power generation group, two wind turbines, two battery storage systems and four groups of loads.
[0107] The maximum charge / discharge power of the energy storage battery system (ESS) in distributed power generation 1 is 250kW, the lower limit of energy is 200kWh, the upper limit is 1000kWh, and the initial energy is 500kWh; the maximum inductive reactive power and the maximum capacitive reactive power of the reactive compensator are both 420kVar; the maximum charge / discharge power of the energy storage battery system 1 (ESS1) is 300kW, the lower limit of energy is 260kWh, the upper limit is 1300kWh, and the initial energy is 650kWh; the maximum charge / discharge power of the energy storage battery system 2 (ESS2) is 150kW, the lower limit of energy is 100kWh, the upper limit is 500kWh, and the initial energy is 250kWh. At the same time, assuming that the system operates at 11:00, the power values of photovoltaic, wind turbine and load are shown in Table 1.
[0108] Table 111: Load demand and renewable energy output at 00:00
[0109]
[0110]
[0111] The other parameters in distributed generation 1 and distributed generation 2 are as follows: The weight coefficients in the objective function are ρ loss =1,ρ rer =1,ρ ess =1; the charge and discharge loss coefficients of the energy storage battery are respectively a ch = 0.1 and a dis =0.15; the charge and discharge efficiency are respectively taken as η c =0.95,η d =0.95; the quadratic coefficient of the abandoned light cost is a r =0.1, the coefficient of the first-order term is b r =1; unit network loss cost is c loss =2$ / kWh.
[0112] In this section, a quadratic programming model of decentralized optimal power flow is built on the Matlab / Simulink platform, and the ode15s solver is called to solve it. The distribution network and the two distributed generation sources are dynamically simulated independently in their respective modules, and the power and voltage information of the boundary nodes are transmitted to the adjacent areas, while receiving the power and voltage information related to the boundary nodes from the adjacent areas.
[0113] Figure 4 The trajectory curves of the voltage amplitude of some nodes (nodes 2, 12, 25 and 33) in the distribution system during the dynamic feedback control process are shown. It can be found that the voltage amplitude of each node gradually increases from 0 to around 1p.u, then begins to oscillate slightly, and finally stabilizes at a certain value. When the trajectory curve of the voltage amplitude tends to a straight line, it can be determined that the voltage amplitude has reached a convergence state.
[0114] Figure 5 and Figure 6 The trajectory curves of the active output power values of each component in distributed power source 1 and distributed power source 2 are given respectively. Similar to the trajectory curve of the voltage amplitude, each power curve first rises or falls sharply, then oscillates slightly, and finally reaches a convergence state. It should be noted that after entering the convergence state, the charging power of the energy storage system in distributed power source 1 and distributed power source 2 is 0, and the discharging power is greater than 0, indicating that at the optimal solution obtained by convergence, the energy storage system will not charge and discharge at the same time, that is, the complementary constraints of charging power and discharging power can be satisfied.
[0115] Figure 7 The boundary interaction power trajectory curves after the distributed power generation is equalized on the distribution network side and the boundary interaction power trajectory curves after the distributed power generation is equalized on the distribution network side are given respectively. It can be seen that for the same boundary, the interaction power on the distribution network side and the interaction power on the distributed power generation side are not the same. With the operation of the feedback control dynamic system, the trajectories of the two will tend to be the same and eventually converge to the same value.
[0116] [Second embodiment]
[0117] See also Figure 8 The present application also provides a multi-distributed power supply and main distribution network optimization operation system 100, the multi-distributed power supply and main distribution network optimization operation method recorded in the above embodiment is applied to the operation system 100, and the operation system 100 includes:
[0118] Distribution network model building module 110: taking network loss as the optimization target and building the optimal distribution network model under the conditions of satisfying linearized AC power flow constraints, node voltage constraints and generator output constraints;
[0119] Distributed power model building module 120: takes network loss cost, conventional generator operation cost, power abandonment cost and energy storage charging and discharging loss cost as optimization targets, and builds the power supply optimal model under the conditions of satisfying linearized AC power flow constraints, node voltage constraints, generator output constraints, renewable energy generation constraints and energy storage system power constraints;
[0120] Planning model construction module 130: using Ward equivalence theory, the network and power information inside the distributed power source is equivalent to an equivalent injection power on the boundary node; after completing the Ward equivalence, the distributed power source transmits the equivalent power as boundary information to the distribution network; and obtains the boundary coupling constraint when the distributed power source is equivalent; using Ward equivalence theory, firstly the network and power information of the distributed power source is equivalent to a node injection power inside the distribution network, and then the network and power information of the distribution network is equivalent to an equivalent injection power on the boundary node; when the distribution network and the distributed power source complete the Ward equivalence, the equivalent power is transmitted to the distributed power source as boundary information; and obtains the boundary coupling constraint when the distribution network is equivalent; construct a quadratic programming model for solving the distributed optimal power flow based on the distribution network optimal model, the power source optimal model and the boundary coupling constraint when performing the equivalence.
[0121] Optimization module 140: constructs an augmented Lagrangian function for solving the quadratic programming model of decentralized optimal power flow, and according to the principle of the saddle point dynamic method, from the perspective of dynamic control stability, converts the KKT condition of an original convex optimization problem into an asymptotic stability process of a dynamic feedback control system, and obtains the optimal solution after solving the KKT condition.
[0122] This operating system has all the technical features of the above-mentioned optimized operating method, which will not be described in detail here.
[0123] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the scope defined by the claims.
Claims
1. A method for optimizing operation based on multiple distributed power sources and main distribution networks. It is characterized in that Optimized operation methods include: Obtain the power flow conditions of the distribution network in the target area, and build an optimal power flow model of the distribution network according to the power flow conditions, which is recorded as the optimal model of the distribution network; Obtain the grid-connected power flow of distributed power sources in the target area, and build an optimal power flow model of distributed power sources based on the grid-connected power flow, which is recorded as the power source optimal model; Through Ward equivalence theory, the original coupling network equations of distribution network and distributed generation are decoupled; A quadratic programming model for solving decentralized optimal power flow is constructed based on the decoupling results. The quadratic programming model is optimized and solved by the saddle point dynamic method.
2. According to the method for optimizing operation based on multiple distributed power sources and main distribution network according to claim 1, It is characterized in that The power flow conditions of the distribution network are obtained in the target area, and the optimal power flow model of the distribution network is constructed according to the power flow conditions, which is recorded as the optimal distribution network model, including: Taking network loss as the optimization target, the optimal distribution network model is constructed under the conditions of satisfying linearized AC power flow constraints, node voltage constraints and generator output constraints.
3. According to the method for optimizing operation based on multiple distributed power sources and main distribution network according to claim 1, It is characterized in that The grid-connected power flow of distributed power sources is obtained in the target area, and the optimal power flow model of distributed power sources is constructed according to the grid-connected power flow, which is recorded as the optimal power model, including: The optimization objectives are network loss costs, conventional generator operating costs, power abandonment costs and energy storage charging and discharging loss costs, and an optimal power supply model is constructed under the conditions of satisfying linearized AC power flow constraints, node voltage constraints, generator output constraints, renewable energy generation constraints and energy storage system power constraints.
4. According to the method for optimizing operation based on multiple distributed power sources and main distribution network according to claim 1, It is characterized in that Through the Ward equivalence theory, the original coupling network equations of the distribution network and distributed generation are decoupled, including: Using the Ward equivalence theory, the network and power information inside the distributed power source is equivalent to an equivalent injected power on the boundary node; after completing the Ward equivalence, the distributed power source transmits the equivalent power as the boundary information to the distribution network; and the boundary coupling constraints of the distributed power source when performing equivalence are obtained.
5. The method for optimizing operation based on multiple distributed power sources and main distribution network according to claim 1, It is characterized in that Through the Ward equivalence theory, the original coupling network equations of the distribution network and distributed generation are decoupled, including: Using the Ward equivalence theory, the network and power information of the distributed power source is firstly equivalent to a node injection power inside the distribution network, and then the network and power information of the distribution network is equivalent to an equivalent injection power on the boundary node; when the distribution network and the distributed power source complete the Ward equivalence, the equivalent power is transmitted to the distributed power source as the boundary information; and the boundary coupling constraints when the distribution network is equivalent are obtained.
6. The method for optimizing operation based on multiple distributed power sources and main distribution network according to claim 1, It is characterized in that According to the decoupling results, a quadratic programming model for solving decentralized optimal power flow is constructed, which includes: A quadratic programming model for solving decentralized optimal power flow is constructed based on the optimal distribution network model, the optimal power source model and the boundary coupling constraints during equivalence.
7. The method for optimizing operation based on multiple distributed power sources and main distribution network according to claim 1, It is characterized in that The quadratic programming model is optimized and solved by the saddle point dynamic method, including: An augmented Lagrangian function is constructed for the quadratic programming model for solving decentralized optimal power flow. According to the principle of the saddle point dynamic method, from the perspective of dynamic control stability, the KKT condition of an original convex optimization problem is converted into an asymptotic stability process of a dynamic feedback control system. After solving the KKT condition, the optimal solution is obtained.
8. An optimized operation system based on multiple distributed power sources and main distribution networks, It is characterized in that The optimized operation method according to any one of claims 1 to 7 is applied to the operation system, and the operation system comprises: Distribution network model building module: takes network loss as the optimization target and builds the optimal distribution network model under the conditions of satisfying linear AC power flow constraints, node voltage constraints and generator output constraints; Distributed power model building module: takes network loss cost, conventional generator operation cost, power abandonment cost and energy storage charging and discharging loss cost as optimization targets, and builds the optimal power model under the conditions of satisfying linearized AC power flow constraints, node voltage constraints, generator output constraints, renewable energy generation constraints and energy storage system power constraints; Planning model construction module: using Ward equivalence theory, the network and power information inside the distributed power source is equivalent to an equivalent injection power on the boundary node; after completing the Ward equivalence, the distributed power source transmits the equivalent power as boundary information to the distribution network; and obtains the boundary coupling constraints when the distributed power source is equivalent; using Ward equivalence theory, firstly the network and power information of the distributed power source is equivalent to a node injection power inside the distribution network, and then the network and power information of the distribution network is equivalent to an equivalent injection power on the boundary node; when the distribution network and the distributed power source complete the Ward equivalence, the equivalent power is transmitted to the distributed power source as boundary information; and obtains the boundary coupling constraints when the distribution network is equivalent; construct a quadratic programming model for solving the decentralized optimal power flow based on the optimal model of the distribution network, the optimal model of the power source and the boundary coupling constraints when performing the equivalence. Optimization module: constructs an augmented Lagrangian function for solving the quadratic programming model of decentralized optimal power flow. According to the principle of saddle point dynamic method, from the perspective of dynamic control stability, the KKT condition of an original convex optimization problem is converted into an asymptotic stability process of a dynamic feedback control system. After solving the KKT condition, the optimal solution is obtained.
9. A device for optimizing operation based on multiple distributed power sources and main distribution networks. It is characterized in that include: Memory for storing computer programs; A processor is used to read and execute a computer program stored in a memory. When the computer program is executed, the processor executes the method for optimizing operation based on multiple distributed power sources and a main distribution network as claimed in any one of claims 1 to 7.
Citation Information
Cited By
User side resource aggregation scheduling method, system and related device
CN120725388A
