CN107681670A - Power network distribution idle work optimization method - Google Patents

Abstract

The invention relates to a reactive power optimization method for power grid distribution, which solves the deficiencies in the prior art. The technical solution is as follows: A, obtain the basic parameters of the target power grid; B, read the reactive power compensation nodes and distribution Network reconfiguration branch; C, determine the optimization objective function of the target power grid; D, select the distribution reactive power compensation node of the power grid, and use the hybrid leapfrog algorithm to optimally select the distribution network reconfiguration branch and perform corresponding actions. The present invention integrates two technologies of reactive power optimization and distribution network reconfiguration, establishes a distribution network comprehensive optimization model with the goal of minimizing the annual comprehensive cost, and adopts a hybrid leapfrog algorithm to solve it, which can not only further reduce network loss and increase node voltage, but also fully Considering the economy of reactive power compensation, the comprehensive optimization of the distribution network is more in line with the economic principle, and the economic value of comprehensive optimization can also be fully tapped.

Description

Reactive power optimization method for power grid distribution

technical field

This patent relates to a power grid construction method, in particular to a reactive power optimization method for power distribution in a power grid.

Background technique

Distribution network reactive power optimization generally uses fixed-point switching reactive power compensation equipment distribution network reconstruction and distribution network reactive power optimization as two important technical means for distribution network optimization operation, which play an important role in ensuring power quality and reducing network loss. . Distribution network reconfiguration obtains the network topology under the optimal optimization target value by changing the closure of the network switch to minimize active power loss and ensure a higher voltage level. In essence, the distribution network reconfiguration is a nonlinear combinatorial optimization problem, and the reactive power optimization of the distribution network is a nonlinear integer programming problem. method of iteration. There are also optimizations based on reconstruction and reactive power optimization respectively, and intelligent algorithms are used to solve them. Compared with the alternate iterative method, the calculation accuracy is improved, but these existing technologies do not carry out reconstruction and reactive power optimization simultaneously in the true sense. , the potential value of comprehensive optimization has not been fully researched and excavated. At the same time, the above studies on the comprehensive optimization of distribution network all take the network loss as the optimization target, and do not consider the economics of reactive power compensation.

Contents of the invention

The purpose of the present invention is to solve the problem that the above-mentioned comprehensive optimization research of distribution network in the prior art takes network loss as the optimization target, and does not consider the economics of reactive power compensation, and provides a reactive power optimization method for power grid distribution.

The technical solution adopted by the present invention to solve the technical problem is: a reactive power optimization method for power grid power distribution, the following steps:

A. Obtain the basic parameters of the target power grid;

B. Read the grid distribution reactive power compensation nodes and distribution network reconstruction branches;

C, determine the optimization objective function of the target grid;

D. Select the grid distribution reactive power compensation node and use the hybrid leapfrog algorithm to perform the optimal selection of distribution network reconfiguration branches and perform corresponding actions.

The present invention integrates two technologies of reactive power optimization and distribution network reconfiguration, establishes a distribution network comprehensive optimization model with the goal of minimizing the annual comprehensive cost, and adopts a hybrid leapfrog algorithm to solve it, which can not only further reduce network loss and increase node voltage, but also fully Considering the economy of reactive power compensation, the comprehensive optimization of the distribution network is more in line with the economic principle, that is, the economic value of comprehensive optimization is fully exploited.

As a preference, in step C, the optimization objective function of the target grid is:

λ is the electricity price; T _max is the annual maximum load loss hours; k ₁ is the annual maintenance cost rate of the compensation equipment; k ₂ is the investment recovery coefficient; Q _i is the reactive power compensation amount of the i-th node, and C ₁ is the reactive power Compensation price; C ₂ is the installation cost of a single compensation point, n is the number of reactive power compensation points;

P _loss is the active power loss of the network, which is the sum of the active power losses of each line, and the expression is:

N _b represents the number of branches; R _k represents the resistance of branch k; S _k represents the closed state of the branch, 1 represents closed, 0 represents open; P _k represents the active power of branch k; Q _k represents the non-active power of branch k. Work power; V _k represents the terminal voltage of branch k;

The constraints are:

V ^min ≤ V _j ≤ V ^max ; _{0≤Qi≤Qi ,max} ;

V ^min represents the upper limit of the node voltage during normal operation of the distribution network;

V ^max represents the lower limit of the node voltage during normal operation of the distribution network;

S ^max represents the maximum carrying capacity of line k;

Q _i,max represents the upper limit of the compensation capacity of the i-th compensation point.

As a preference, the leapfrog algorithm parameters are set as follows: for a power grid with a total active load range of 3000kW to 4000kW and a total reactive load range of 2000kvar to 2500kvar, the frog group size is set to 80, the number of groups is 20, and the number of global evolutions is 50 , the number of local evolutions is 3; the amount of reactive compensation is 1.2 times the total reactive load of the system as the upper limit of reactive compensation, the minimum search step is set manually, λ = 0.5 yuan/kW·h, T _max = 5000h, k ₁ =0.13, k ₂ =0.1, C ₁ =60 yuan/kvar, C ₂ =5000 yuan/node.

Preferably, the minimum search step size is 10kvar.

As a preference, all nodes with the heaviest reactive power load in the power grid are selected as reactive power compensation points.

As a preference, when calculating the active power loss P _loss of the network, set several branches as the change regulation branch. In the middle of the road, after changing the value of S _k in the previously changed adjustment branch, re-execute the hybrid leapfrog algorithm for the optimal selection of the distribution network reconstruction branch; repeat the execution in turn until the hybrid leapfrog algorithm performs the distribution network reconstruction branch The optimal choice satisfies the constraints.

Preferably, the change regulation branch is sorted according to the total active load or total reactive load of the distribution network reconfiguration branch, and the sorting order is the smaller the total active load or the total reactive power load of the distribution network reconfiguration branch The higher the priority, if the hybrid leapfrog algorithm fails to meet the constraints when optimally selecting the distribution network reconfiguration branch, the distribution network reconfiguration branch with higher priority will be changed first.

Preferably, the change regulation branch is sorted according to the total active load or total reactive load of the distribution network reconfiguration branch, and the sorting order is the active total load or reactive power of the distribution network reconfiguration branch with _Sk value=1 The smaller the total load, the higher the sorting priority. The sorting order is that the active total load or reactive total load of the distribution network reconstruction branch with S _k value = 0 is smaller, and the sorting priority is lower. If the hybrid leapfrog algorithm performs When the optimal selection of the distribution network reconfiguration branch cannot meet the constraint conditions, the distribution network reconfiguration branch with high priority should be changed first.

The substantive effect of the present invention is: the present invention integrates the two technologies of reactive power optimization and distribution network reconfiguration, establishes a distribution network comprehensive optimization model with the goal of minimizing the annual comprehensive cost, and adopts the hybrid leapfrog algorithm to solve it, which can not only further reduce the network It also fully considers the economics of reactive power compensation, making the comprehensive optimization of the distribution network more in line with the economic principle, and can fully tap the economic value of comprehensive optimization.

detailed description

The technical solution of the present invention will be further specifically described below through specific examples.

Example 1:

A reactive power optimization method for power grid distribution, the following steps:

A. Obtain the basic parameters of the target power grid;

B. Read the grid distribution reactive power compensation nodes and distribution network reconstruction branches;

C, determine the optimization objective function of the target grid;

D. Select the grid distribution reactive power compensation node and use the hybrid leapfrog algorithm to perform the optimal selection of distribution network reconfiguration branches and perform corresponding actions.

In step C, the optimization objective function of the target power grid is:

λ is the electricity price; T _max is the annual maximum load loss hours; k ₁ is the annual maintenance cost rate of the compensation equipment; k ₂ is the investment recovery coefficient; Q _i is the reactive power compensation amount of the i-th node, and C ₁ is the reactive power Compensation price; C ₂ is the installation cost of a single compensation point, n is the number of reactive power compensation points;

P _loss is the active power loss of the network, which is the sum of the active power losses of each line, and the expression is:

N _b represents the number of branches; R _k represents the resistance of branch k; S _k represents the closed state of the branch, 1 represents closed, 0 represents open; P _k represents the active power of branch k; Q _k represents the non-active power of branch k. Work power; V _k represents the terminal voltage of branch k;

The constraints are:

V ^min ≤ V _j ≤ V ^max ; _{0≤Qi≤Qi ,max} ;

V ^min represents the upper limit of the node voltage during normal operation of the distribution network;

V ^max represents the lower limit of the node voltage during normal operation of the distribution network;

S ^max represents the maximum carrying capacity of line k;

Q _i,max represents the upper limit of the compensation capacity of the i-th compensation point.

Leapfrog algorithm parameters are set as follows: For a power grid with a total active load range of 3000kW to 4000kW and a total reactive load range of 2000kvar to 2500kvar, the size of the frog group is set to 80, the number of groups is 20, the number of global evolutions is 50, and the number of local evolutions is 50. The number of times is 3; the amount of reactive power compensation is 1.2 times the total reactive load of the system as the upper limit of reactive power compensation, the minimum search step is set manually, λ = 0.5 yuan/kW·h, T _max = 5000h, k ₁ = 0.13, k ₂ =0.1, C ₁ =60 yuan/kvar, C ₂ =5000 yuan/node.

The minimum search step size is 10kvar. Select all the nodes with the heaviest reactive power load in the power grid as reactive power compensation points. When calculating the active power loss P _loss of the network, set several branches as the change regulation branch. If the optimal selection of the distribution network reconfiguration branch by the hybrid leapfrog algorithm cannot meet the constraint conditions, in the change regulation branch, After changing the value of S _k in the previously changed adjustment branch, re-execute the hybrid leapfrog algorithm for the optimal selection of the distribution network reconstruction branch; repeat the execution in turn until the hybrid leapfrog algorithm performs the optimal selection of the distribution network reconstruction branch. Constraints are met when selecting. When calculating the active power loss P _loss of the network, set several branches as the change regulation branch. If the optimal selection of the distribution network reconfiguration branch by the hybrid leapfrog algorithm cannot meet the constraint conditions, in the change regulation branch, After changing the value of S _k in the previously changed adjustment branch, re-execute the hybrid leapfrog algorithm for the optimal selection of the distribution network reconstruction branch; repeat the execution in turn until the hybrid leapfrog algorithm performs the optimal selection of the distribution network reconstruction branch. Constraints are met when selecting.

Example 2:

This embodiment is basically the same as Embodiment 1, the difference is that the change regulation branch is sorted according to the total active load or total reactive load of the distribution network reconstruction branch, and the sorting order is the distribution with S _k value = 1 The smaller the total active load or the total reactive load of the network reconfiguration branch, the higher the sorting priority, and the smaller the total active load or total reactive load of the distribution network reconfiguration branch with the value of S _k = 0, the higher the sorting order is The lower the sorting priority, if the hybrid leapfrog algorithm fails to meet the constraints when optimally selecting the distribution network reconfiguration branch, the distribution network reconfiguration branch with higher priority will be changed first. In this embodiment, the hybrid leapfrog algorithm includes the following steps:

Step 1: Initialize parameters, including: the number F of the frog group; the number m of the group; the number n of frogs in the group; the maximum _{allowable jumping} step S _max ; Difference solution P _w ; global iterative evolution times N _g , local iterative evolution times N ₁ , upper limit Q _{i, max} of reactive power compensation at each compensation point;

Step 2: Randomly generate the initial frog group, and calculate the evaluation value of each frog;

Step 3: Sort in ascending order according to the evaluation value, record the optimal solution P _z , and divide the frog population into groups in the following way: put the first frog into the first group, put the second frog into the second group , put the mth frog into the mth group, put the m+1th frog into the first group, and so on, until all the frogs are put into the designated position; step 4: carry out each group according to the following formula evolutionary operation

S=ceil(Rand()×(P _w -P _b ));

NewP _w =P _w +S, -S _min ≤ S ≤ S _max ;

Among them, ceil means rounding, rand() means to generate a random number from 0 to 1, S means the leapfrog step size, S _max and S _min are the leapfrog step length limits, and NewP _w means the updated P _w ;

Step 5: After all groups are updated, calculate the evaluation value of all frogs in the group;

Step 6: Determine whether the stop condition is met; if so, stop the search, otherwise go to step 3;

Step 7: Execute corresponding actions according to the optimal choice.

The specific steps of step 4 include the following sub-steps:

Step 4-1: Let I _M = _IN =0, I _M represents the counter of group evolution, and _IN represents the counter of local evolution;

Step 4-2: Select P _b and P _w of the current group, and add 1 to I _M ;

Step 4-3: Add 1 to I _N ;

Step 4-4: According to

S=ceil(Rand()×(P _w -P _b ));

NewP _w ＝P _w +S, -S _min ≤ S ≤ S _max Improve the worst frog in the group;

Step 4-5: If the worst frog was improved in the last step, replace the worst frog with this new frog, otherwise replace P _b in formula (8) with P _z and re-evolve;

Step 4-6: If the worst frog has not been improved in the previous step, randomly generate a feasible solution to replace the worst frog;

Step 4-7: If I _N is less than the number of local evolutions L _N , go to step 4-3;

Step 4-8: If I _M is smaller than the group number m, then go to step 4-2, otherwise go to step 5 of the global search.

The present invention integrates two technologies of reactive power optimization and distribution network reconfiguration, establishes a distribution network comprehensive optimization model with the goal of minimizing the annual comprehensive cost, and adopts a hybrid leapfrog algorithm to solve it, which can not only further reduce network loss and increase node voltage, but also fully Considering the economy of reactive power compensation, the comprehensive optimization of the distribution network is more in line with the economic principle, and the economic value of comprehensive optimization can also be fully tapped.

The embodiment described above is only a preferred solution of the present invention, and does not limit the present invention in any form. There are other variations and modifications on the premise of not exceeding the technical solution described in the claims.

Claims (8)

1. A kind of 1. power network distribution idle work optimization method, it is characterised in that following steps：

   A, obtain the basic parameter of target grid；

   B, read power network distribution candidate compensation buses and Distribution system branch road；

   C, determine the optimization object function of target grid；

   D, select power network distribution candidate compensation buses to carry out the optimal selection of Distribution system branch road using shuffled frog leaping algorithm and hold Row corresponding actions.
2. 2. power network distribution idle work optimization method according to claim 1, it is characterised in that in step C, target grid Optimization object function is：

   <mrow> <mi>min</mi> <mi> </mi> <mi>F</mi> <mo>=</mo> <msub> <mi>&amp;lambda;T</mi> <mi>max</mi> </msub> <msub> <mi>P</mi> <mrow> <mi>l</mi> <mi>o</mi> <mi>s</mi> <mi>s</mi> </mrow> </msub> <mo>+</mo> <mrow> <mo>(</mo> <msub> <mi>k</mi> <mn>1</mn> </msub> <mo>+</mo> <msub> <mi>k</mi> <mn>2</mn> </msub> <mo>)</mo> </mrow> <mrow> <mo>(</mo> <msub> <mi>C</mi> <mn>1</mn> </msub> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>n</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msub> <mi>Q</mi> <mi>i</mi> </msub> <mo>+</mo> <msub> <mi>nC</mi> <mn>2</mn> </msub> <mo>)</mo> </mrow> <mo>;</mo> </mrow>

   λ is electricity price；T_maxHourage is lost for annual peak load；k₁To compensate the year maintenance cost rate of equipment；k₂Reclaimed for investment Coefficient；Q_iFor the reactive-load compensation amount of i-th of node, C₁For the price of reactive-load compensation；C₂For the mounting cost of single compensation point, n is Reactive-load compensation point number；

   P_lossFor the active loss of network, for the summation of every circuit active loss, expression formula is：

   <mrow> <msub> <mi>P</mi> <mrow> <mi>l</mi> <mi>o</mi> <mi>s</mi> <mi>s</mi> </mrow> </msub> <mo>=</mo> <munderover> <mi>&amp;Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <msub> <mi>N</mi> <mi>b</mi> </msub> </munderover> <msub> <mi>S</mi> <mi>k</mi> </msub> <msub> <mi>R</mi> <mi>k</mi> </msub> <mfrac> <mrow> <msup> <msub> <mi>P</mi> <mi>k</mi> </msub> <mn>2</mn> </msup> <mo>+</mo> <msup> <msub> <mi>Q</mi> <mi>k</mi> </msub> <mn>2</mn> </msup> </mrow> <mrow> <msup> <msub> <mi>V</mi> <mi>k</mi> </msub> <mn>2</mn> </msup> </mrow> </mfrac> </mrow>

   N_bRepresent circuitry number；R_kRepresent branch road k resistance；S_kBranch road closure state is represented, 1 represents closure, and 0 represents to open；P_kTable Show branch road k active power；Q_kRepresent branch road k reactive power；V_kRepresent branch road k terminal voltage；

   Constraints is：

   V^min≤V_j≤V^max；0≤Q_i≤Q_{I, max}；

   V^minRepresent the upper limit of node voltage during distribution normal operation；

   V^maxRepresent the lower limit of node voltage during distribution normal operation；

   S^maxRepresent circuit k maximum carrying capacity；

   Q_{I, max}Represent i-th of compensation point compensation capacity upper limit.
3. 3. power network distribution idle work optimization method according to claim 1, it is characterised in that the algorithm parameter that leapfrogs is set such as Under：To active total load scope in 3000kW to 4000kW, idle total load scope 2000kvar to 2500kvar power network, Frog group's size is set as 80, and group's number is 20, and global evolution number is 50, and Local Evolution number is 3；Reactive-load compensation amount is with system 1.2 times of upper limits as reactive-load compensation of total load or burden without work, minimum step-size in search are set manually, λ=0.5 yuan/kWh, T_max =5000h, k₁=0.13, k₂=0.1, C₁=60 yuan/kvar, C₂=5000 yuan/node.
4. 4. power network distribution idle work optimization method according to claim 3, it is characterised in that minimum step-size in search is 10kvar。
5. 5. power network distribution idle work optimization method according to claim 3, it is characterised in that all idle negative in selection power network The most heavy node of lotus is as reactive-load compensation point.
6. 6. power network distribution idle work optimization method according to claim 3, it is characterised in that in the active loss of calculating network P_lossWhen set some branch roads as change adjust branch road, if shuffled frog leaping algorithm carry out Distribution system branch road optimal selection Shi Wufa meets constraints, in change adjusts branch road, by the S in the branch road of change regulation before_kRe-started after value change Shuffled frog leaping algorithm carries out the optimal selection of Distribution system branch road；It is repeated in performing, until shuffled frog leaping algorithm carries out distribution Meet constraints during the optimal selection for reconstructing branch road.
7. 7. power network distribution idle work optimization method according to claim 6, it is characterised in that it is described change regulation branch road according to The active total load or idle total load of Distribution system branch road are ranked up, and clooating sequence is the active total negative of Distribution system branch road Lotus or the smaller then Sort Priority of idle total load are higher, if shuffled frog leaping algorithm carries out the optimal selection of Distribution system branch road It can not meet that constraints then first changes the high Distribution system branch road of priority.
8. 8. power network distribution idle work optimization method according to claim 6, it is characterised in that it is described change regulation branch road according to The active total load or idle total load of Distribution system branch road are ranked up, clooating sequence S_kThe Distribution system branch road of value=1 Active total load or the smaller then Sort Priority of idle total load it is higher, clooating sequence S_kThe Distribution system branch road of value=0 Active total load or idle total load smaller then Sort Priority it is lower, if shuffled frog leaping algorithm carries out Distribution system branch road It can not meet that constraints then first changes priority high Distribution system branch road during optimal selection.