Probabilistic Planning Method and Device of Distributed Generation for Reconfiguration Oriented to Bearing Capacity and Network Loss

Abstract

The present invention discloses a probabilistic planning method and device for distributed power sources oriented to bearing capacity and network loss reconstruction. By solving a two-layer planning model, a distributed power source access scheme and a network reconstruction scheme are determined. The planning layer of the two-layer planning model operates a distributed power source planning model with the maximum access capacity of distributed power sources and the minimum network loss as the objectives, and the operation layer operates a network reconstruction optimization model with the minimum network loss as the objective. The method includes: solving the distributed power source planning model to obtain a distributed power source access scheme; solving the network reconstruction optimization model to obtain a network reconstruction scheme; taking the access location and capacity of the distributed power sources obtained by the planning layer as the input of the operation layer, and taking the network reconstruction scheme and network loss obtained by the operation layer as the input of the planning layer, and performing two-layer optimization iteration until an optimal distributed power source access scheme and an optimal network reconstruction scheme are obtained. The present invention effectively reduces network loss while improving the utilization rate of distributed power sources.

Description

Technical Field

[0001] The present invention relates to the technical field of distributed power planning, and particularly relates to a probabilistic planning method and device for distributed power supply oriented to bearing capacity and network loss reconstruction. Background Art

[0002] In recent years, distributed power sources based on renewable energy have developed rapidly in the distribution network. However, due to the characteristics of non-dispatchable and intermittent power generation of high-penetration distributed power sources, problems such as over-limit of the network end voltage will occur, threatening the safe and stable operation of the distribution network. Moreover, an unreasonable access scheme for distributed power sources will further aggravate adverse effects such as over-limit of voltage. In order to improve the acceptance capacity of distributed power sources, it is necessary to first plan an access scheme considering the uncertainties of load and distributed power source output.

[0003] To further improve the acceptance capacity of distributed power sources, most current studies adopt the method of network reconstruction. By adjusting the network topology structure, the power flow can be effectively improved, which is a powerful way to improve the acceptance capacity of distributed power sources in the distribution network. However, only aiming at maximizing the access capacity of distributed power sources to adjust the network topology structure, with a large number of distributed power sources incorporated into the distribution network, the distribution network shows characteristics such as bidirectional power flow and complex control, and the network loss may increase. An unreasonable topology structure may further increase the network loss. Therefore, when optimizing the network reconstruction scheme, it is necessary to comprehensively consider the acceptance capacity of distributed power sources and the network loss.

[0004] For example, in the related art, the invention patent with the application number 201410116331.1 discloses an intelligent substation load automatic distribution method based on the access of distributed power sources, establishing a mathematical model of distributed power sources, performing power flow calculation for the whole network according to the network and load data collected by the intelligent substation where the distributed power sources are connected to the distribution network, and establishing a reactive power optimization model and a network reconstruction model for the access of distributed power sources to the distribution network under given constraint conditions, and cross-iterating the two sub-problems of the reactive power optimization model and the network reconstruction model to gradually approach the optimal solution to complete the intelligent substation load automatic distribution for the access of distributed power sources to the distribution network. However, this scheme takes the access of distributed power sources as the boundary condition, that is, optimizing reactive power optimization and network reconstruction after the access of distributed power sources to improve the safety and economy of system operation. Instead, it is not for the decision-making work on the maximum acceptance capacity and access scheme of distributed power sources with the distribution network as the main body. Summary of the Invention

[0005] The technical problem to be solved by the present invention is how to effectively reduce the network loss while improving the utilization rate of distributed power sources.

[0006] The present invention realizes the solution to the above technical problem through the following technical means:

[0007] On the one hand, the present invention proposes a probabilistic planning method for distributed power sources oriented to bearing capacity and network loss reconstruction, which is used to solve a two-layer planning model to determine a distributed power source access scheme and a network reconstruction scheme. The two-layer planning model includes an upper planning layer and a lower operation layer. The upper planning layer runs a distributed power source planning model with the maximum distributed power source access capacity and the minimum network loss as the objectives, and the lower operation layer runs a network reconstruction optimization model with the minimum network loss as the objective. The method includes:

[0008] Solve the distributed power source planning model running in the upper planning layer to obtain a distributed power source access scheme;

[0009] Solve the network reconstruction optimization model running in the lower operation layer to obtain a network reconstruction scheme;

[0010] Take the distributed power source access location and capacity obtained by the upper planning layer as the input of the lower operation layer, and take the network reconstruction scheme and network loss obtained by the lower operation layer as the input of the upper planning layer, and perform two-layer optimization iteration until the optimal distributed power source access scheme and the optimal network reconstruction scheme are obtained.

[0011] In the two-layer planning model adopted by the present invention, there is a distributed power source planning model with the maximum distributed power source access capacity and the minimum network loss as the objectives. Its optimization variables are the access location and access capacity of the distributed power source, mainly solving the site selection and capacity determination problem under the maximum distributed power source access capacity; the lower operation layer runs a network reconstruction optimization model with the minimum network loss as the objective, and its optimization variables are the opening and closing states of the tie switches and sectionalizing switches, considering the operation problem; the planning layer transmits the distributed power source access scheme including the distributed power source access location and capacity information to the operation layer, and the operation layer feeds back the network loss to the planning layer. The models running in the upper and lower layers perform two-layer optimization iteration until the final distributed power source access scheme and network reconstruction scheme results are obtained. It can realize the decision-making of the distributed power source access scheme and network reconstruction scheme with the maximum distributed power source acceptance capacity and the minimum network loss under the condition that the indexes of the distribution network do not exceed the limit, which is of great significance for in-depth research on the improvement of the distributed power source acceptance capacity.

[0012] Further, the distributed power source planning model includes a first objective function and a second objective function. The first objective function aims at the maximum distributed power source access capacity, and the second objective function aims at the minimum active network loss. Among them, the formula of the first objective function is as follows:

[0013]

[0014] Among them, f1 is the total access capacity of the distributed power source; P i DG$P_{i}$ is the installed capacity of distributed power at node $i$, which is used as the maximum power generation of the distributed photovoltaic power generation curve in the calculation; $N$ dg is the set of nodes where distributed power can potentially be installed.

[0015] The formula of the second objective function is expressed as follows:

[0016]

[0017] where $P$ loss is the network loss, $N$ is the number of nodes in the distribution network, $b$ ij is the open - circuit status of branch $ij$. When $b$ ij = 1, it means that branch $ij$ is closed; when $b$ ij = 0, it means that branch $ij$ is open. $R$ ij and $X$ ij are the impedance parameters of branch $ij$. $U$ i , $P$ i and $Q$ i are the voltage, active power and reactive power at node $i$ respectively. $U$ j , $P$ j and $Q$ j are the voltage, active power and reactive power at node $j$ respectively. $\theta$ ij is the voltage phase - angle difference between nodes $i$ and $j$.

[0018] Furthermore, the network reconfiguration optimization model aims to minimize the network loss, and its formula is expressed as follows:

[0019]

[0020] where $N$ l is the number of branches in the distribution network, $b$ l is the open - circuit status of branch $l$. When $b$ l = 1, it means that branch $l$ is closed; when $b$ l = 0, it means that branch $l$ is open. $r$ l is the resistance parameter of branch $l$. $U$ i , $P$ i and $Q$ i are the voltage, active power and reactive power at the head - end node $i$ of branch $l$ respectively.

[0021] Furthermore, the constraint conditions of the network reconfiguration optimization model include power - flow equation constraints, node - voltage constraints, line - current - carrying capacity constraints, short - circuit current constraints, network - topology constraints, distributed - power - node installed - capacity constraints, distributed - power siting constraints, distributed - power total - installation - amount constraints, switch - status constraints and switch - switching - times constraints;

[0022] The distributed - power siting constraint is:

[0023] i ∈ N dg

[0024] Among them, N dg is the set of nodes where distributed power sources can potentially be installed;

[0025] The total installed capacity constraint of the distributed power source is:

[0026]

[0027] Among them, α dg is the proportionality coefficient of the distributed photovoltaic grid connection to the total load, N is the number of nodes in the distribution network, P i load is the load of the i-th node, P i DG is the installed capacity of the distributed power source at node i.

[0028] Furthermore, the method further includes:

[0029] Using a fuzzy set to process the uncertain variables included in the bilevel programming model to obtain an uncertainty model, and the representation form of the fuzzy set is:

[0030]

[0031]

[0032] Among them, P i load is the active power load at node i, Δ P i load and are respectively the load prediction value, the allowed downward deviation and the allowed upward deviation at node i, P i dg is the photovoltaic active power injected at node i, Δ P i dg and are respectively the distributed power source output prediction value, the allowed downward deviation and the allowed upward deviation at node i.

[0033] Furthermore, the method further includes:

[0034] Representing the uncertain variables using standardized deviation variables as follows:

[0035]

[0036]

[0037]

[0038] Among them, are the downward standardized deviation of load, the upward standardized deviation of load, the downward standardized deviation of distributed power source, and the upward standardized deviation of distributed power source respectively;

[0039] Among them, satisfies the following constraint conditions:

[0040]

[0041] Among them, Γ is the degree of conservatism.

[0042] Furthermore, solving the distributed power source planning model operating in the upper planning layer to obtain a distributed power source access plan includes:

[0043] Using the robust equality to form a min-max two-layer model outside the two-layer planning model;

[0044] Normalizing the distributed power source planning model to obtain the comprehensive objective function of the upper planning layer;

[0045] Using the decomposition-based multi-objective evolutionary method MOEAD to optimize and solve the comprehensive objective function to obtain a distributed power source access plan in the form of a Pareto solution set;

[0046] Using the fuzzy theory to sort the Pareto solution set to obtain a sorted distributed power source access plan.

[0047] Furthermore, solving the network reconfiguration optimization model operating in the lower operation layer to obtain a network reconfiguration plan includes:

[0048] Using a genetic algorithm to solve the network reconfiguration optimization model to obtain the network reconfiguration plan, where the encoding process uses a decimal circular encoding strategy to encode each branch, and the fitness value is the network loss.

[0049] On the other hand, the present invention also proposes a distributed power source probabilistic planning device for bearing capacity and network loss reconstruction, which is used to solve the two-layer planning model to determine a distributed power source access plan and a network reconfiguration plan. The two-layer planning model includes an upper planning layer and a lower operation layer. The upper planning layer operates a distributed power source planning model with the maximum access capacity of the distributed power source and the minimum network loss as the goals, and the lower operation layer operates a network reconfiguration optimization model with the minimum network loss as the goal. The device includes:

[0050] The first solution module is used to solve the distributed power source planning model operating in the upper-layer planning layer to obtain a distributed power source access scheme;

[0051] The second solution module is used to solve the network reconfiguration optimization model operating in the lower-layer operation layer to obtain a network reconfiguration scheme;

[0052] The optimization iteration module is used to take the access location and capacity of the distributed power source obtained from the upper-layer planning layer as the input of the lower-layer operation layer, and take the network reconfiguration scheme and network loss obtained from the lower-layer operation layer as the input of the upper-layer planning layer, and perform two-layer optimization iteration until the optimal distributed power source access scheme and the optimal network reconfiguration scheme are obtained.

[0053] Furthermore, the distributed power source planning model includes a first objective function and a second objective function. The first objective function aims to maximize the access capacity of the distributed power source, and the second objective function aims to minimize the active power loss. Among them, the formula of the first objective function is expressed as follows:

[0054]

[0055] Among them, f1 is the total access capacity of the distributed power source; P i DG is the installed capacity of the distributed power source at node i, and it is used as the maximum power generation of the distributed photovoltaic power generation curve in the calculation; N dg is the set of nodes where the distributed power source can potentially be installed.

[0056] The formula of the second objective function is expressed as follows:

[0057]

[0058] Among them, P loss is the power loss, N is the number of nodes in the distribution network, b ij is the open / close state of branch ij. When b ij = 1, it means that branch ij is closed. When b ij = 0, it means that branch ij is open. R ij , X ij are the impedance parameters of branch ij, U i , P i and Q i are the voltage, active power and reactive power at node i respectively, U j , P j and Q j are the voltage, active power and reactive power at node j respectively, θ ij is the voltage phase angle difference between nodes i and j;

[0059] The network reconstruction optimization model aims to minimize network losses, and is expressed by the following formula:

[0060]

[0061] Where N l is the number of branches in the distribution network, b l is the opening and closing state of branch l. When b l = 1, it means that branch l is closed. When b l = 0, it means that branch l is disconnected. r l is the resistance parameter of branch l, and U i , P i and Q i are the voltage, active power, and reactive power at the head node i of branch l, respectively.

[0062] The advantages of the present invention are as follows:

[0063] (1) In the two-layer programming model adopted by the present invention, there is a distributed power source planning model with the maximum access capacity of distributed power sources and the minimum network losses as the objectives. Its optimization variables are the access location and access capacity of distributed power sources, mainly solving the problem of site selection and capacity determination under the maximum access capacity of distributed power sources; the lower-layer operation layer operates a network reconstruction optimization model with the minimum network losses as the objective, and its optimization variables are the opening and closing states of tie switches and sectionalizing switches, considering the operation problem; the planning layer transmits the distributed power source access plan, including the access location and access capacity information of distributed power sources, to the operation layer, and the operation layer feeds back the network losses to the planning layer. The models operating in the upper and lower layers perform two-layer optimization iterations until the final distributed power source access plan and network reconstruction plan results are obtained. It can realize the decision-making of the distributed power source access plan and network reconstruction plan with the maximum distributed power source acceptance capacity and the minimum network losses under the condition that the indicators of the distribution network do not exceed the limits. While improving the utilization rate of distributed power sources, it effectively reduces network losses, which is of great significance for in-depth research on the improvement of the distributed power source acceptance capacity.

[0064] (2) In order to improve the acceptance capacity of distributed power sources, the present invention plans to consider the access plan with uncertainties in load and distributed power source output, making the planning results more in line with the actual situation and more reasonable.

[0065] (3) The genetic algorithm is used to solve the network reconstruction optimization model in the operation layer, and the decimal circular coding strategy is adopted for coding each branch during the coding process of the genetic algorithm, and elitist retention is carried out to reduce the solution space of distribution network reconstruction.

[0066] Additional aspects and advantages of the present invention will be given in part in the following description, become apparent in part from the following description, or be understood through the practice of the present invention. Description of the Drawings

[0067] Figure 1 is the flowchart of the distributed power source probabilistic planning method for bearing capacity and network loss reconstruction in the first embodiment of the present invention;

[0068] Figure 2 is the structural diagram of the bi-level programming model in the first embodiment of the present invention;

[0069] Figure 3 is the overall flowchart of the distributed power source probabilistic planning method for bearing capacity and network loss reconstruction in the first embodiment of the present invention;

[0070] Figure 4 is the structural diagram of the distributed power source probabilistic planning device for bearing capacity and network loss reconstruction in the second embodiment of the present invention. Detailed implementation manners

[0071] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts fall within the scope of protection of the present invention.

[0072] As Figures 1 to 2 shown, the embodiments of the present invention propose a distributed power source probabilistic planning method for bearing capacity and network loss reconstruction, which is used to solve the bi-level programming model to determine the distributed power source access scheme and the network reconstruction scheme. The bi-level programming model includes an upper planning layer and a lower operation layer. The upper planning layer runs a distributed power source planning model with the maximum distributed power source access capacity and the minimum network loss as the objectives, and the lower operation layer runs a network reconstruction optimization model with the minimum network loss as the objective. The method includes the following steps:

[0073] S10. Solve the distributed power source planning model running in the upper planning layer to obtain a distributed power source access scheme;

[0074] S20. Solve the network reconstruction optimization model running in the lower operation layer to obtain a network reconstruction scheme;

[0075] S30. Use the distributed power source access location and capacity obtained by the upper planning layer as the input of the lower operation layer, and use the network reconstruction scheme and network loss obtained by the lower operation layer as the input of the upper planning layer to perform bi-level optimization iteration until the optimal distributed power source access scheme and the optimal network reconstruction scheme are obtained.

[0076] It should be noted that the upper-layer planning layer aims to maximize the access capacity of distributed power sources and minimize network losses. Its optimization variables are the access locations and capacities of distributed power sources, mainly solving the problem of site selection and capacity determination under the maximum access capacity of distributed power sources. The lower-layer operation layer operates with the goal of minimizing network losses. Its optimization variables are the opening and closing states of tie switches and sectional switches, considering operation issues. The planning layer transmits the distributed power source access plan, including distributed power source access location and capacity information, to the operation layer. The operation layer feeds back the network losses to the planning layer. The models of the upper and lower layers perform two-layer optimization iterations until the final distributed power source access plan and network reconstruction plan results are obtained. It is possible to achieve the decision-making of the distributed power source access plan and network reconstruction plan with the goals of maximizing the acceptance capacity of distributed power sources and minimizing network losses under the condition that the indicators of the distribution network do not exceed the limits. While improving the utilization rate of distributed power sources, it effectively reduces network losses, which is of great significance for in-depth research on the improvement of the acceptance capacity of distributed power sources.

[0077] In one embodiment, the distributed power source planning model includes a first objective function and a second objective function. The first objective function aims to maximize the access capacity of distributed power sources, and the second objective function aims to minimize the active network losses. Among them, the formula of the first objective function is expressed as follows:

[0078]

[0079] Among them, f1 is the total access capacity of distributed power sources; P i DG is the installed capacity of the distributed power source at node i, which is used as the maximum power generation power of the distributed photovoltaic power generation curve in the calculation; N dg is the set of nodes where distributed power sources can potentially be installed.

[0080] When the distributed power source is connected to the distribution network, the active power loss of the distribution network is related to the access location and access capacity of the distributed power source. The total active network loss of the distribution network, that is, the formula of the second objective function, is expressed as follows:

[0081]

[0082] Among them, P loss is the network loss, N is the number of nodes in the distribution network, b ij is the opening and closing state of branch ij. When b ij =1, it means that branch ij is closed. When b ij =0, it means that branch ij is open. R ij , X ij are the impedance parameters of branch ij, U i , P i and Q i are the voltage, active power, and reactive power at node i respectively, Uj , P j and Q j are the voltage, active power, and reactive power at node j, respectively, and θ ij is the voltage phase angle difference between nodes i and j.

[0083] It should be noted that since a large number of distributed power sources are integrated into the distribution network, the distribution network shows characteristics such as bidirectional power flow and complex control, and the network loss may increase. In this embodiment, a multi-objective function is set at the planning layer, and the objectives include: the maximum access capacity of distributed power sources and the minimum active network loss, so as to obtain a distributed power source siting and sizing scheme with improved distributed power source acceptance ability and reduced grid connection loss.

[0084] In one embodiment, the network reconfiguration optimization model aims to minimize the network loss, and the formula is expressed as follows:

[0085]

[0086] where N l is the number of branches in the distribution network, b l is the open / close state of branch l. When b l = 1, it means that branch l is closed; when b l = 0, it means that branch l is open. r l is the resistance parameter of branch l, and U i , P i and Q i are the voltage, active power, and reactive power at the head node i of branch l, respectively.

[0087] It should be noted that since the network loss required in the planning layer needs to be obtained through power flow calculation. Therefore, this embodiment sets an operation layer, and uses network reconfiguration optimization to improve the distributed power source acceptance ability and change the network topology to reduce losses. Among them, the power flow is calculated on the network topology after considering the network reconfiguration scheme.

[0088] In one embodiment, the constraint conditions of the network reconfiguration optimization model include power flow equality constraints, node voltage constraints, line current-carrying capacity constraints, short-circuit current constraints, network topology constraints, distributed power source node installation capacity constraints, distributed power source siting constraints, distributed power source total installation constraints, switch state constraints, and switch switching times constraints, where:

[0089] (1) Power flow equality constraint conditions

[0090]

[0091] In the formula, P i , Q i are the active injection power and reactive injection power at node i; Ui , U j is the voltage amplitude of nodes i and j; G ij , B ij are the conductance and susceptance of branch ij; θ ij is the voltage phase angle difference between nodes i and j; P i G , Q i G are respectively the active power and reactive power injected from the superior power grid at node i; P i dg , Q i dg are respectively the active power and reactive power of the photovoltaic power injected at node i; P i load , Q i load are respectively the active load and reactive load at node i.

[0092] (2) Node voltage constraint conditions

[0093] U imin < U i < U imax

[0094] In the formula, U imin is the lower limit value of the node voltage, taken as 0.9, U imax is the upper limit value of the node voltage, taken as 1.07.

[0095] (3) Line current-carrying capacity constraint conditions

[0096] The power flowing through the line is limited by its maximum apparent power:

[0097]

[0098] In the formula, P l and Ql are respectively the active power and reactive power flowing through line l, S lmax is the maximum apparent power that line l is allowed to carry.

[0099] (4) Short-circuit current constraint conditions

[0100] The connection of the power source to the power grid will increase the short-circuit current level of the power grid. After the connection of the distributed power source, the short-circuit current of the line should not exceed the limit value of this voltage level:

[0101] I SCL < I SCLmax

[0102] In the formula, I SCL and I SCLmax are respectively the maximum short-circuit current of the line and the maximum short-circuit current of the circuit breaker.

[0103] (5) Network topology constraint conditions

[0104] Before and after reconstruction, the distribution network structure should be radial without loops and islands.

[0105] (6) Installation capacity constraint conditions for distributed power generation nodes

[0106] Installation capacity constraint for distributed power generation nodes:

[0107] 0 ≤ P i DG ≤ P DGmax

[0108] In the formula, P DGmax is the maximum limit of the access of distributed power generation in the line. If the installation capacity is 0, it means that no distributed power generation is connected to this node.

[0109] (7) Location constraint conditions for distributed power generation

[0110] Restricted by external factors such as geography and load level, distributed power generation cannot be installed at all nodes. To improve the rationality and availability of the planning scheme, the installation location needs to be optimized within the given location set, that is, i ∈ N dg .

[0111] It should be noted that due to external factors such as geography and load level, distributed power generation cannot be installed at all nodes. To improve the rationality and availability of the planning scheme, the installation location needs to be optimized within the given location set. Therefore, the location constraint conditions for distributed power generation are set in this embodiment.

[0112] (8) Total installation amount constraint conditions for distributed power generation

[0113]

[0114] In the formula, α dg is the proportionality coefficient between the grid connection of distributed photovoltaics and the total load, N is the number of distribution network nodes, P load i is the load of the i-th node, and P i DG is the installed capacity of distributed power generation at node i.

[0115] (9) Switch state constraint and switch switching times constraint conditions

[0116] The branch opening state is affected by the states of the tie switch and the sectionalizing switch:

[0117]

[0118] The initial state of the tie switch is open, and the initial state of the sectionalizing switch is closed:

[0119]

[0120] In the formula, X ij ins is the opening and closing state of the tie switch on branch ij, 1 represents the closed state, and 0 represents the open state; X ij ses is the opening and closing state of the sectionalizing switch on branch ij, 1 represents the closed state, and 0 represents the open state.

[0121] Considering the switch switching loss, a constraint on the number of switch switches is imposed:

[0122]

[0123]

[0124] In the formula, S max is the maximum number of switch switches within a day.

[0125] It should be noted that since network reconfiguration requires switching the switch state and considering the switch switching loss during the switching, a constraint on the number of switch switches is set in this embodiment.

[0126] (10) Distributed power output constraint

[0127] P i dg = η i P i DG

[0128] In the formula, η i is the distributed power output coefficient of node i.

[0129] In one embodiment, as Figure 3 shown, the method further includes:

[0130] Using a fuzzy set to process the uncertain variables included in the bilevel programming model to obtain an uncertainty model, and the representation form of the fuzzy set is:

[0131]

[0132]

[0133] Among them, P i load is the active load at node i, Δ P i load and are the load prediction value, the allowable downward deviation, and the allowable upward deviation at node i, respectively, P i dg is the active power injected by the superior power grid at node i, Δ P i dg 、 are the predicted output value of the distributed power source, the allowable downward deviation, and the allowable upward deviation at node i, respectively.

[0134] It should be noted that distributed power sources based on renewable energy have developed rapidly in the distribution network. However, the characteristics of non-dispatchable and intermittent power generation of high-penetration distributed power sources will bring problems such as over-limit of the voltage at the end of the network, threatening the safe and stable operation of the distribution network. At the same time, unreasonable selection of the access scheme for distributed power sources will further aggravate adverse effects such as voltage over-limit. In order to improve the acceptance capacity of distributed power sources, the access scheme considering the uncertainty of load and distributed power source output is planned in this embodiment, making the planning result more in line with the actual situation and more reasonable.

[0135] In one embodiment, the method further includes:

[0136] The uncertainty variables are represented by using standardized deviation variables as follows:

[0137]

[0138]

[0139]

[0140] Among them, are the load downward standardized deviation amount, the load upward standardized deviation amount, the distributed power source downward standardized deviation amount, and the distributed power source upward standardized deviation amount, respectively;

[0141] Among them, satisfy the following constraint conditions:

[0142]

[0143] Among them, Γ is the degree of conservatism.

[0144] In this embodiment, the uncertainty variables are directly represented by using standardized deviation variables, and by adjusting the degree of conservatism, the disadvantage of the robust optimization being too conservative is overcome.

[0145] In one embodiment, step S10 includes the following steps:

[0146] S11. Use a robust peer-to-peer approach to form a min-max two-layer model outside the two-layer programming model;

[0147] It should be noted that robust optimization is used to solve the uncertainties of distributed power generation output and load. A robust peer-to-peer equation with uncertainty conservatism constraints is established through the conservatism parameter, and a min-max two-layer optimization model is formed outside the original planning layer - operation layer two-layer model to solve the robust peer-to-peer equation.

[0148] S12. Normalize the distributed power generation planning model to obtain the comprehensive objective function of the upper planning layer;

[0149] Specifically, take the reciprocal of the first objective function f1, that is, f1 = 1 / f1. After that, minimize the second objective function f2 and the reciprocal of f1, that is, minimize the total objective function F. Normalize f2 and the reciprocal of f1 and construct the objective function. Its mathematical expression is as follows:

[0150]

[0151]

[0152] In the formula, f i is the true value of the i-th objective function, f imax is the maximum value of the i-th objective function, f imin is the minimum value of the i-th objective function, ω1 and ω2 are the weight coefficients of the two objective functions respectively, ω1 + ω2 = 1, and are the functions after normalizing the objective functions f1 and f2 respectively.

[0153] Then the comprehensive objective function of the planning layer can be expressed as:

[0154] minF(f1,f2).

[0155] S13. Use the decomposition-based multi-objective evolutionary method MOEAD to optimize and solve the comprehensive objective function to obtain a distributed power generation access scheme in the form of a Pareto solution set;

[0156] It should be noted that in this embodiment, the multi-objective evolutionary method MOEAD based on decomposition is used to solve the multi-objective distributed power generation planning layer considering the maximum access capacity and the minimum network loss. First, the population is initialized, and uniformly distributed weight vectors are generated as reference points. The Chebyshev method is used for decomposition to obtain sub-problems corresponding to the two objective functions. Then, mating and mutation are performed on the two objectives respectively. Simulated binary crossover and polynomial mutation are used for mutation to obtain a new generation of population. Then, individuals that do not meet the constraint conditions are replaced and corrected with individuals within the boundaries. Then, the reference point z is updated according to the aggregation function values of the two objectives, and z takes the current z and the one with the minimum aggregation function value. Then, the weight neighborhood and EP are updated, and the dominated vectors are removed from the EP. Until the set maximum number of iterations is reached, the loop ends, and the Pareto optimal solution set of the distributed power generation access scheme is obtained.

[0157] S14. Use fuzzy theory to rank the Pareto solution set to obtain the ranked distributed power generation access scheme.

[0158] It should be noted that in this embodiment, fuzzy theory is used to rank the obtained Pareto solution set in order to select the distributed power generation access scheme:

[0159]

[0160] where Γ j is the fuzzy membership degree of the j-th optimal solution in the Pareto optimal solution set; N i is the number of objective functions, which is taken as 2 here; N j is the number of optimal solutions in the Pareto optimal solution set; M j,i is the membership degree value of the i-th objective function in the j-th Pareto optimal solution, and is calculated as follows:

[0161]

[0162] where M i is the membership degree value of the i-th objective function; f i is the true numerical value of the i-th objective function; is the function after normalization of the i-th objective function; f imax is the maximum value of the i-th objective function, and f imin is the minimum value of the i-th objective function.

[0163] In one embodiment, the step S20 includes the following steps:

[0164] The genetic algorithm is used to solve the network reconfiguration optimization model to obtain the network reconfiguration scheme. Among them, the encoding process uses the decimal circular encoding strategy to encode each branch, and the fitness value is the network loss.

[0165] It should be noted that the basic operation process of the genetic algorithm is as follows: initialization, encoding, calculating fitness to evaluate individuals, selection, crossover, and mutation. In this embodiment, the population is initialized to generate N network reconstruction schemes; in order to reduce the solution space of the distribution network reconstruction, a decimal circular encoding strategy is adopted to encode each branch; the network loss is used as the fitness value, and elitist retention, selection, crossover, and mutation are performed to obtain the optimal network reconstruction scheme.

[0166] In this embodiment, the reasonable layout of distributed power sources is realized by optimizing the access scheme of distributed power sources, and at the same time, the purpose of improving the acceptance capacity of distributed power sources and improving the network loss is achieved by optimizing the network reconstruction scheme.

[0167] In addition, as Figure 4 shown, an embodiment of the present invention also proposes a probability planning device for distributed power sources for bearing capacity and network loss reconstruction, which is used to solve the bilevel programming model to determine the distributed power source access scheme and the network reconstruction scheme. The bilevel programming model includes an upper planning layer and a lower operation layer. The upper planning layer runs a distributed power source planning model with the maximum access capacity of distributed power sources and the minimum network loss as the objectives, and the lower operation layer runs a network reconstruction optimization model with the minimum network loss as the objective. The device includes:

[0168] The first solving module 10 is used to solve the distributed power source planning model running in the upper planning layer to obtain the distributed power source access scheme;

[0169] The second solving module 20 is used to solve the network reconstruction optimization model running in the lower operation layer to obtain the network reconstruction scheme;

[0170] The optimization iteration module 30 is used to use the distributed power source access location and capacity obtained by the upper planning layer as the input of the lower operation layer, and use the network reconstruction scheme and network loss obtained by the lower operation layer as the input of the upper planning layer to perform bilevel optimization iteration until the optimal distributed power source access scheme and the optimal network reconstruction scheme are obtained.

[0171] In one embodiment, the distributed power source planning model includes a first objective function and a second objective function. The first objective function aims at the maximum access capacity of distributed power sources, and the second objective function aims at the minimum active network loss. Among them, the formula of the first objective function is expressed as follows:

[0172]

[0173] Among them, f1 is the total access capacity of distributed power sources; P i DG$P_{i}$ is the installed capacity of the distributed power source at node $i$, which is used as the maximum power generation of the distributed photovoltaic power generation curve in the calculation; $N$ dg is the set of nodes where distributed power sources can potentially be installed.

[0174] The formula of the second objective function is expressed as follows:

[0175]

[0176] Among them, $P$ loss is the network loss, $N$ is the number of nodes in the distribution network, $b$ ij is the switching state of branch $ij$. When $b$ ij = 1, it means that branch $ij$ is closed; when $b$ ij = 0, it means that branch $ij$ is open. $R$ ij , $X$ ij are the impedance parameters of branch $ij$. $U$ i , $P$ i and $Q$ i are the voltage, active power, and reactive power at node $i$ respectively. $U$ j , $P$ j and $Q$ j are the voltage, active power, and reactive power at node $j$ respectively. $\theta$ ij is the voltage phase angle difference between nodes $i$ and $j$;

[0177] The network reconfiguration optimization model aims to minimize the network loss, and its formula is expressed as follows:

[0178]

[0179] Among them, $N$ l is the number of branches in the distribution network, $b$ l is the switching state of branch $l$. When $b$ l = 1, it means that branch $l$ is closed; when $b$ l = 0, it means that branch $l$ is open. $r$ l is the resistance parameter of branch $l$. $U$ i , $P$ i and $Q$ i are the voltage, active power, and reactive power at the head node $i$ of branch $l$ respectively.

[0180] In one embodiment, the constraint conditions of the network reconfiguration optimization model include power flow equation constraints, node voltage constraints, line current-carrying capacity constraints, short-circuit current constraints, network topology constraints, distributed power source node installation capacity constraints, distributed power source location constraints, total distributed power source installation constraints, switch state constraints, and switch switching times constraints.

[0181] In one embodiment, the device further includes:

[0182] A processing module, which is used to process the uncertainty variables included in the bilevel programming model by using a fuzzy set to obtain an uncertainty model. The representation form of the fuzzy set is:

[0183]

[0184]

[0185] where P i load is the active power load at node i, Δ P i load , are respectively the load prediction value, the allowable downward deviation, and the allowable upward deviation at node i. P i dg is the active power injected from the superior power grid at node i, Δ P i dg , are respectively the predicted value of the distributed power generation output, the allowable downward deviation, and the allowable upward deviation at node i.

[0186] Specifically, in this embodiment, the uncertainty variables are also represented by using standardized deviation variables as follows:

[0187]

[0188]

[0189]

[0190] where are respectively the load downward standardized deviation quantity, the load upward standardized deviation quantity, the distributed power generation downward standardized deviation quantity, and the distributed power generation upward standardized deviation quantity;

[0191] where satisfies the following constraint conditions:

[0192]

[0193] where Γ is the conservatism.

[0194] In one embodiment, the first solving module 10 includes:

[0195] A conversion unit, which is used to form a min-max bilevel model outside the bilevel programming model by using a robust equality.

[0196] A normalization unit, configured to normalize the distributed power planning model to obtain the comprehensive objective function of the upper-layer planning layer;

[0197] A solving unit, configured to use the multi-objective evolutionary algorithm based on decomposition MOEAD to optimize and solve the comprehensive objective function to obtain a distributed power access scheme in the form of a Pareto solution set;

[0198] A sorting unit, configured to sort the Pareto solution set using fuzzy theory to obtain a sorted distributed power access scheme.

[0199] In one embodiment, the second solving module 20 is specifically configured to:

[0200] Solve the network reconfiguration optimization model using a genetic algorithm to obtain the network reconfiguration scheme, where in the encoding process, a decimal circular encoding strategy is used to encode each branch, and the fitness value is the network loss.

[0201] It should be noted that other embodiments or implementation methods of the distributed power probability planning device for bearing capacity and network loss reconstruction according to the present invention can refer to the above method embodiments, and will not be repeated here.

[0202] It should be noted that the logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then storing it in a computer memory.

[0203] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logical functions on data signals, application specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0204] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0205] In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" can explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "a plurality" is at least two, such as two, three, etc., unless otherwise specifically and clearly defined.

[0206] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A probabilistic planning method for distributed power sources oriented to bearing capacity and network loss reconstruction, characterized in that Used to solve the bilevel programming model to determine the distributed power generation access scheme and network reconfiguration scheme. The bilevel programming model includes an upper planning layer and a lower operation layer. The upper planning layer runs a distributed power generation planning model with the goals of maximizing the access capacity of distributed power generation and minimizing network losses. The lower operation layer runs a network reconfiguration optimization model with the goal of minimizing network losses. The method includes: Solving the distributed power generation planning model running in the upper planning layer to obtain the distributed power generation access scheme, specifically including forming a min-max bilevel model outside the bilevel programming model using robust duality; normalizing the distributed power generation planning model to obtain the comprehensive objective function of the upper planning layer; using the decomposition-based multi-objective evolutionary method MOEAD to optimize and solve the comprehensive objective function to obtain the distributed power generation access scheme in the form of a Pareto solution set; using fuzzy theory to rank the Pareto solution set to obtain the ranked distributed power generation access scheme; Solving the network reconfiguration optimization model running in the lower operation layer to obtain the network reconfiguration scheme, specifically including using a genetic algorithm to solve the network reconfiguration optimization model to obtain the network reconfiguration scheme. Among them, the encoding process uses a decimal circular encoding strategy to encode each branch, and the fitness value is the network loss; Taking the distributed power generation access location and capacity obtained by the upper planning layer as the input of the lower operation layer, and taking the network reconfiguration scheme and network loss obtained by the lower operation layer as the input of the upper planning layer, and performing bilevel optimization iteration until the optimal distributed power generation access scheme and the optimal network reconfiguration scheme are obtained.

2. The probabilistic planning method for distributed power sources oriented to bearing capacity and network loss reconstruction according to claim 1, wherein The distributed power generation planning model includes a first objective function and a second objective function. The first objective function aims to maximize the access capacity of distributed power generation, and the second objective function aims to minimize the active network loss. Among them, the formula of the first objective function is expressed as follows: Among them, f 1 is the total access capacity of distributed power sources; P i DG is the installed capacity of the distributed power source at node i , which is used as the maximum power generation of the distributed photovoltaic power generation curve in the calculation; N dg is the set of nodes where distributed power sources can potentially be installed; The formula of the second objective function is expressed as follows: Among them, is the network loss, N is the number of distribution network nodes, b ij is the branch ij open - circuit status. When b ij = 1, it means the branch ij is closed. When b ij = 0, it means the branch ij is open. R ij and X ij are the impedance parameters of the branch ij . U i and P i and Q i are the voltage, active power and reactive power at the node i respectively. U j and P j and Q j are the voltage, active power and reactive power at the node j respectively. is the voltage phase - angle difference between the nodes i and j .

3. The probabilistic planning method for distributed power sources oriented to bearing capacity and network loss reconstruction according to claim 1, wherein, The network reconfiguration optimization model aims to minimize the network loss, and the formula is expressed as follows: Among them, N l is the number of distribution network branches, b l is the l opening and closing state of the branch. When b l = 1, it means the branch l is closed. When b l = 0, it means the branch l is open. r l is the resistance parameter of the branch l . U i , P i and Q i are respectively the voltage, active power and reactive power at the head node l of the branch i .

4. The probabilistic planning method for distributed power sources oriented to bearing capacity and network loss reconstruction according to claim 2, characterized in that, The constraint conditions of the network reconfiguration optimization model include power flow equation constraints, node voltage constraints, line current-carrying capacity constraints, short-circuit current constraints, network topology constraints, distributed power generation node installation capacity constraints, distributed power generation site selection constraints, total distributed power generation installation constraints, switch state constraints, and switch switching times constraints; The distributed power generation site selection constraint is: Among them, N dg is the set of nodes that potentially can install distributed power sources; The total distributed power generation installation constraint is: Among them, α dg is the proportionality coefficient of the distributed photovoltaic grid connection and the total load,[[]] N is the node set of the research feeder,[[]] P i load is the i load of the th node,[[]] i is the installed capacity of the distributed power source at node 5. The probabilistic planning method for distributed power sources oriented to bearing capacity and network loss reconstruction according to any one of claims 1-4, characterized in that The method further includes: Using a fuzzy set to process the uncertain variables included in the bilevel programming model to obtain an uncertain model. The representation form of the fuzzy set is: Among them, P i load is the active power load at the node i ; are respectively the load prediction value, the allowable downward deviation, and the allowable upward deviation at the node i ; is the active power of photovoltaic power injected at the node i ; are respectively the predicted value of the distributed power generation output, the allowable downward deviation, and the allowable upward deviation at the node i .

6. The probabilistic planning method for distributed power sources oriented to bearing capacity and network loss reconstruction according to claim 5, wherein The method further includes: Using a standardized deviation variable to represent the uncertain variable as follows: Among them, are the downward standardized deviation of load, the upward standardized deviation of load, the downward standardized deviation of distributed power source, and the upward standardized deviation of distributed power source, respectively; Among them, satisfy the following constraint conditions: where Γ is the degree of conservatism.

7. A distributed power source probabilistic planning device for bearing capacity and network loss reconstruction, characterized in that For solving the bi-level programming model to determine the distributed power source access scheme and the network reconfiguration scheme, the bi-level programming model includes an upper planning layer and a lower operation layer. The upper planning layer runs a distributed power source planning model with the maximum distributed power source access capacity and the minimum network loss as the objectives, and the lower operation layer runs a network reconfiguration optimization model with the minimum network loss as the objective. The device includes: A first solving module for solving the distributed power source planning model running in the upper planning layer to obtain a distributed power source access scheme, specifically including forming a min-max bi-level model outside the bi-level programming model by using robust equality; normalizing the distributed power source planning model to obtain the comprehensive objective function of the upper planning layer; using the multi-objective evolutionary algorithm based on decomposition MOEAD to optimize and solve the comprehensive objective function to obtain a distributed power source access scheme in the form of a Pareto solution set; using the fuzzy theory to sort the Pareto solution set to obtain a sorted distributed power source access scheme; A second solving module for solving the network reconfiguration optimization model running in the lower operation layer to obtain a network reconfiguration scheme, specifically including using a genetic algorithm to solve the network reconfiguration optimization model to obtain the network reconfiguration scheme, where the encoding process encodes each branch using a decimal circular encoding strategy, and the fitness value is the network loss; An optimization iteration module for using the distributed power source access location and capacity obtained by the upper planning layer as the input of the lower operation layer, and using the network reconfiguration scheme and network loss obtained by the lower operation layer as the input of the upper planning layer to perform bi-level optimization iteration until an optimal distributed power source access scheme and an optimal network reconfiguration scheme are obtained.

8. The distributed power source probabilistic planning device for bearing capacity and network loss reconstruction according to claim 7, characterized in that, The distributed power source planning model includes a first objective function and a second objective function. The first objective function aims at the maximum distributed power source access capacity, and the second objective function aims at the minimum active network loss. Among them, the formula of the first objective function is expressed as follows: Among them, f 1 is the total access capacity of distributed power sources; P i DG is the installed capacity of the distributed power source at node i , which is used as the maximum power generation of the distributed photovoltaic power generation curve in the calculation; N dg is the set of nodes where distributed power sources can potentially be installed; The formula of the second objective function is expressed as follows: Among them, is the network loss, N is the number of nodes in the distribution network, b ij is the branch ij open - circuit state. When b ij = 1, it means that the branch ij is closed. When b ij = 0, it means that the branch ij is open. R ij , X ij are the impedance parameters of the branch ij . U i , P i and Q i are the voltage, active power and reactive power at the node i respectively. U j , P j and Q j are the voltage, active power and reactive power at the node j respectively. is the voltage phase - angle difference between the nodes i , j . The network reconfiguration optimization model aims at the minimum network loss, and the formula is expressed as follows: Among them, N l is the number of branches in the distribution network, b l is the l opening and closing state of the branch. When b l = 1, it means the branch l is closed. When b l = 0, it means the branch l is open. r l is the resistance parameter of the branch l . U i , P i and Q i are respectively the voltage, active power and reactive power at the head node l of the branch i .

Images