CN105071433B - A kind of configuration scheme of distributed generation resource - Google Patents
A kind of configuration scheme of distributed generation resource Download PDFInfo
- Publication number
- CN105071433B CN105071433B CN201510464144.7A CN201510464144A CN105071433B CN 105071433 B CN105071433 B CN 105071433B CN 201510464144 A CN201510464144 A CN 201510464144A CN 105071433 B CN105071433 B CN 105071433B
- Authority
- CN
- China
- Prior art keywords
- power supply
- optimization
- minimum
- distributed power
- objective
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a kind of configuration scheme of distributed generation resource, angle of the program present invention from planning, by the interests for considering distribution operator, meter and the randomness of wind-power electricity generation, establish that active energy loss by power distribution network is minimum, total voltage deviation is minimum and the multiple-objection optimization allocation models of risk cost minimum composition herein；The normalization of multiple target is realized using fuzzy set theory, there is no each different problem of sub-goal dimension；To the Optimal Allocation Model established herein, solved using TSP question particle cluster algorithm, because which introducing mutation operation, so that the situation that basic particle group algorithm is easily trapped into local optimum is improved, finally, what it is as example is 33 Node power distribution systems of IEEE, and the AMPSO for demonstrating the Optimal Allocation Model of DG proposed in this paper and selecting, obtained simulation result shows the model used herein and algorithm is feasible validity.
Description
Technical Field
The invention relates to an optimal configuration scheme of a distributed power supply, and belongs to the technical field of power systems.
Background
Distributed power sources are increasingly connected into a power distribution network for grid-connected operation due to the characteristics of high efficiency, small investment scale, network loss reduction, flexible power generation mode, various energy types, environmental protection and the like. The distributed power supply can be installed in a load center so as to track changes of loads timely, is more economical in power utilization peak periods compared with centralized power supply, and can save power transmission and transformation investment, stabilize voltage, reduce energy consumption, and improve power quality, power supply reliability, flexibility and safety of operation of a power system when the distributed power supply and the centralized power supply are operated in a combined mode.
The method is characterized in that the position of the DG connected into the power distribution network and the capacity of the connected electric energy have certain influence on the planning and the operation of the DG, such as the influence on the system voltage distribution, the voltage stability, the magnitude of short-circuit current, the operation state, relay protection and the like, the reasonable determination of the DG connection position and the DG connection electric energy capacity has very important effect on the improvement of the benefits of the power distribution network and the safe and stable operation, according to the characteristics of the power distribution network, the scientific and reasonable DG connection position and the access electric energy capacity are searched under the condition of meeting the related technical constraint conditions of the power distribution network, and the main content of DG optimal configuration is to reduce the influence on the normal operation of the power distribution network caused by the DG connection as much as possible.
However, the conventional distributed power supply optimization configuration generally only considers a single aspect of economy or technology, cannot comprehensively consider a plurality of targets, and has certain limitation.
Disclosure of Invention
The purpose of the invention is: aiming at the defects of the prior art, an optimal configuration scheme of a distributed power supply is provided, and a new method is provided for the existing distributed power supply to be connected to a power distribution network so as to overcome the defects of the prior art.
Technical scheme of the invention
An optimal configuration scheme of a distributed power supply is characterized in that the optimal configuration of the distributed power supply is researched from the planning angle by considering the benefits of a power distribution operator and the randomness of the power generation output of the distributed power supply, namely, the internet access position and the capacity of the distributed power supply are solved, a multi-objective optimal configuration model consisting of the minimum active power loss, the minimum total voltage deviation and the minimum risk cost of a power distribution network is established, and constraint conditions are set and the optimal configuration model is optimized; the multi-objective normalization is realized by adopting a fuzzy set theory, the relationship among all sub-objectives is well coordinated, and the effect of overall optimization is further achieved; and solving by using a self-adaptive variation particle swarm algorithm.
In the above optimized configuration scheme for a distributed power supply, the multi-objective optimized configuration model specifically includes the following steps:
1. the active electric energy loss is minimum;
the distribution network voltage grade is lower, the R/X value is larger, the network loss is larger in load flow calculation, and the network loss can be reduced by reasonably accessing a distributed power supply, so that the minimum network loss is taken as a target function to bring better economic benefit, and since the minimum network loss is taken as an optimization target from the planning angle, the target function with the minimum active energy loss is as follows:
in the formula: I.C. A i Current for branch i; n is the total branch number of the system; r is i Is the branch resistance, and h is the hours of the programming period.
2. The optimization target model with the minimum total voltage deviation is as follows;
the out-of-limit of the node voltage amplitude value can affect the normal work of a user and the safety of a system, so that the node voltage can be closer to a reference value by taking the total node voltage deviation as a target function, the voltage level of a power distribution network is guaranteed, and the effect of improving the voltage quality is achieved. The objective function for minimum total voltage deviation is:
in the formula: detV is the total voltage deviation; n is the number of network nodes; u shape b Is a reference voltage; u shape i Is the voltage value of the ith node.
3. The risk cost is minimum;
the distributed power supplies of renewable energy sources such as wind power generation, photovoltaic power generation and the like are easily affected by natural environment or other factors, so that the output of the distributed power supplies has the characteristics of randomness, volatility and uncertainty, the randomness of the output of the distributed power supplies is reflected through risk cost, when the generated energy of the distributed power supplies cannot reach an expected value, the cost representation of power supply is carried out through a power grid, namely a superior power supply, and the objective function of the risk cost is as follows:
C risk ＝P·C·t·n
in the formula: p is the probability that the DG (namely, the distributed power supply) with randomness outputs less than the expected output; c is the cost of power supply by the superior power supply when the DG (namely the distributed power supply) output is less than the expected output; t is the annual running time of the DG (i.e., the distributed power supply); and n is the planning year.
In the above optimized configuration scheme for the distributed power supply, the constraint conditions of the multi-objective optimized configuration model are inequality constraints, and include node voltage constraints, distributed power supply access constraints, and branch current constraints.
1. Node voltage constraint
According to the technical principle of power grid operation in China, the upper and lower voltage limits of a node in a 10kV power grid are required to be between 1.07p.u. and 0.93p.u., and after a distributed power supply is connected, local voltage is possibly out of limit, so that the node voltage is ensured to meet corresponding limiting conditions, and the node voltage constraint is as follows:
in the formula:the lowest and highest voltages of node i, respectively, and N is the number of network nodes.
2. Capacity constraints for distributed power access to electrical energy
Distributed generator does not all receive the electric wire netting dispatch because of its power output with open and stop, can cause very big impact to the user when the distribution network inserts the distributed generator of too big capacity, in order to control because of distributed generator's the influence of access to the distribution network, need put down the capacity that distributed generator inserts, specifically as follows:
∑S DG ≤η∑S LD
in the formula: s DG The total capacity of distributed power sources for accessing a power distribution network; s LD Is the total system load; eta is the upper limit of the proportion of the total capacity of the distributed power supply to the total load of the system, and the value of the eta is 0.4.
3. Current confinement
In the formula:and (4) allowing the passing of an upper current limit for the ith branch.
The equality constraints are the nodal power flow equations:
in the formula: p is i Injecting active power for the node i; q i Injecting reactive power for node i; g ij Is the conductance between nodes i and j; b is ij Is the susceptance between nodes i and j; u shape i 、U j The voltage amplitudes of node i and node j, respectively.
In the above-mentioned optimized configuration scheme of a distributed power supply, the method for implementing multi-objective normalization by using a fuzzy set theory first performs maximum processing on the minimum value of the membership of the multi-objective to improve the poor index, thereby achieving the purpose of improving the overall performance of the system, while the adopted membership function is a piecewise linear function, for the optimization containing a plurality of objective functions, the optimization objectives are often contradictory and difficult to simultaneously achieve the optimization, and under the condition that the importance of each optimization objective is the same, the maximum satisfaction method is adopted to solve, that is, the larger the minimum value of the membership function of the optimization objective is, the larger the overall satisfaction is.
The specific multi-target normalization method comprises the following steps: due to the fact that the active power loss, the total voltage deviation and the risk cost are different in dimension, when the weight coefficient method is adopted, the weight coefficient is not easy to determine, the result obtained through optimization is probably obtained by making up for a poor index through a good index, and the effect of overall optimization cannot be achieved. Therefore, the fuzzy set theory is adopted to realize multi-target normalization, and the minimum value of the membership degree is subjected to maximization treatment to improve the index of difference, so that the aim of improving the overall performance of the system is fulfilled.
1. Membership function for optimization objectives
The requirement of the decision maker on the objective function can be embodied by the difference of the membership function, and the common membership function mainly comprises: piecewise linear functions, power exponentials, hyperbolas, linear and inverse hyperbolas, and the like. For the above functions, linear and piecewise linear functions are more common in comparison, mainly because inverse hyperbola, power exponent, etc. will increase the nonlinear characteristics of the decision making process, making the solution more difficult, and here, piecewise linear function is selected as membership function to optimize each sub-target, the membership function is as shown in fig. 1, and the specific function form is as follows:
in the formula: mu.s i As an objective function f i Is taken as the membership degree of [0,1 ]]When the membership degree is 1, the decision maker is completely satisfied with the optimization result; when the membership degree is 0, the decision maker is completely unsatisfied with the optimization result;is the expected value of the objective function;is the limit value of the objective function.
2. Multi-objective normalization
For the optimization including a plurality of objective functions, the respective optimization objectives are often contradictory and difficult to achieve the optimization at the same time, so that generally, in practical applications, it is only required that the respective optimization objectives achieve the optimal values as much as possible. In the case where the importance of each optimization objective is the same, the maximum satisfaction method [62] is usually used to solve. The overall satisfaction is defined as:
M＝min{μ 1 ,μ 2 ,μ 3 }
from the above formula, it can be seen that the larger the minimum value of the membership function of the optimization objective is, the larger the overall satisfaction is, and no matter how the membership function values of the other two optimization objectives change, the membership function values of the two optimization objectives must not be smaller than the minimum membership function value, so that the overall performance can be improved by increasing the minimum membership function value, and the whole can be optimized. Then, under the condition of satisfying the corresponding constraint conditions, the original multi-objective optimization problem is converted into the problem of maximizing the overall satisfaction degree M value:
maxM＝min{μ 1 ,μ 2 ,μ 3 }。
in the above optimal configuration scheme for the distributed power source, the optimal configuration algorithm for the distributed power source is solved by using a self-adaptive variation particle swarm algorithm, which specifically includes:
1. basic particle swarm optimization
Assuming that the particle swarm algorithm searches optimally in an N-dimensional space, the algorithm will generate a group of random particles after initialization, each particle has a corresponding N-dimensional vector X i ＝(x 1 ,x 2 ,x 3 ......,x iN ) And V i ＝(v 1 ,v 2 ,v 3 ......,v iN ) The position and velocity indicated. Each particle can judge the quality of the particle by comparing fitness function values obtained through calculation, and evolves by continuously changing the position and the speed of the particle until a global optimal solution is found out. The optimum position of the particle found in each iteration is denoted p best The best position searched by the whole population is recorded as g best 。
For each iteration, the velocity and position of the particles in the population are updated by the following equations:
(4-1)
in the formula: n represents the nth dimension of the N-dimensional solution space, N = l,2,3, \8230;, N; k is the number of iterations; omega is an inertia weight and is a non-negative constant; r is a radical of hydrogen 1 And r 2 Is in [0,1 ]]A random number generated in between; c. C 1 And c 2 For acceleration factors, typically c 1 ＝c 2 ＝2。
As can be seen from the formula (4-1), the particlesThe speed variation of (2) is divided into three blocks: the former partThe influence of the speed of the particle at the last moment on the current speed is represented, and the particle has the balance capability of local and global searching, namely motion inertia; the second partThe self-thinking of the particles in the searching process is shown, the particles are guided to move towards the optimal position of the individual, and the self-thinking part of the particles is shown; third partThe social recognition part of the particles represents the sharing of information among the particles, and guides the particles to move toward the optimal position of the population. Through the three parts, the particles continuously change the positions of the particles according to self experiences and population experiences.
The inertia weight omega has a very important role as a parameter in the PSO, mainly characterizes the influence capacity of the inertia of the particle on the speed, and is introduced into the global and local searching capacity capable of controlling the algorithm. The larger the value of omega is, the larger the search space of the particles becomes, and the particles are likely to expand to a new region for searching, so that the global search capability is stronger; the smaller the omega value is, the stronger the local searching capability of the algorithm is, and more detailed searching can be carried out around the currently searched solution; when ω is 0, the particles lose their memory function, the current velocity will be independent of the last moment velocity, and the update of the particle position is determined only by the positions of the individual extreme and global optima.
Therefore, omega is subjected to self-adaptive linear decrement along with the progress of the iterative process of the algorithm, so that the convergence performance of the algorithm is obviously improved, the inertial weight can be well controlled by adopting a linear decrement weight, and omega of the kth iteration is as follows:
(4-2)
in the formula: omega min And omega max Respectively, minimum and maximum inertia weight; and N is the maximum iteration number.
The particle swarm algorithm mainly follows the following five basic principles:
l) proximity principle (proximity): the particle swarm must be capable of simple spatial and temporal calculations;
2) Quality principle (quality): the particle swarm must be able to react to the quality factors of the surrounding environment;
3) Diversity principle (diverseresP 0 nse): the particle population should not behave within an overly narrow range;
4) Qualitative principle (stability): the particle swarm should not change its own behavior every time the environment changes;
5) Adaptability principle (adaptability): with an acceptable amount of computation, the population needs to be able to change their behavior at the appropriate time.
The basic particle swarm algorithm steps are as follows:
1) The position and velocity of each particle in the population is randomly generated in an N-dimensional space.
2) The fitness function value of each particle is calculated according to the fitness function, at p best In which the current position and the fitness function value of each particle are stored, in g best The position and fitness function value of the optimal particle in the population are stored.
3) The velocity and position of each particle is updated according to equation (4-1).
4) The fitness value of each particle is compared with the p of the particle best By comparison, if it is better than p best Then replace p with the current particle value best Otherwise, keeping the state unchanged; comparing the current p best And the population optimum value g best By comparison, if it is better than g best Then use p best Substituted g best Otherwise, the value remains unchanged.
5) Stopping searching if the convergence condition is met, and outputting the global optimal solution, otherwise, turning to the step 2).
2. Self-adaptive variation particle swarm algorithm
If the algorithm is trapped in local optimum, mutation operation can be carried out to search other spaces to find the optimum solution, and new p can be found possibly in other spaces best And g best And circulating until a global optimal solution is found, and judging the convergence degree of an Adaptive Mutation Particle Swarm Algorithm (AMPSO) by using the fitness variance of the population on the basis of the idea, thereby adaptively changing the Mutation probability and carrying out Mutation operation on a global extreme value by using a random disturbance method.
(I) group fitness variance
The fitness function value of each particle is determined by its location, and the degree of aggregation of each particle may be reflected by the overall change in the fitness function value of all particles. When the fitness function values of the particles are close to unity, then the population fitness variance tends to 0. Defining a fitness variance δ of a population 2 Comprises the following steps:
(4-3)
in the formula: n is the population scale; f. of k Is the fitness value of particle k; f is a normalized scaling factor which limits delta 2 Is determined by the following formula:
f＝max{max(|f k -f a ),1}
(4-4)
f a mean fitness of the population:
(4-5)
as can be seen from the formula (4-4), the particlesThe degree of clustering of the subgroups is determined by the fitness variance δ of the population 2 In response to the magnitude of, delta 2 The smaller the value is, the more the population tends to converge; delta 2 The larger the value, the more random the population is still searching.
(II) mutation operation
When the fitness variance delta of the population 2 When the particle swarm is small, the particle swarm may fall into local optimum, and if the self-adaptive variation operation is performed according to the global optimum solution at the moment, the capability of the particle to jump out of the local optimum solution is enhanced, in other words, the variation probability should follow the adaptability variance delta 2 The magnitude of the value varies. The mutation probability is determined by the following formula:
P k ＝(P max -P min )(δ 2 /n) 2 +(P min -P max )(2δ 2 /n)+P max
(4-6)
in the formula: p max 、P min The upper and lower limits of the mutation probability. As can be seen from formulas 4 to 6, δ 2 The smaller, P k The larger.
Selecting random disturbance mode to pair g best Performing mutation operation, wherein alpha is a random variable conforming to Guass (0, 1) distribution, and generating a random number r epsilon [0,1 ]]When r is<P k Gaussian mutation was performed:
g best '＝g best (1+0.5α)
(4-7)。
3. the distributed power supply optimization configuration solving method based on the AMPSO comprises the following steps:
1) Initializing; various parametric information input to the power distribution system, including line parameters and load power data, determine corresponding constraints. Initializing PSO parameters including population size m, particle dimension n, acceleration factor c 1 And c 2 Minimum and maximum values of inertial weight ω min And omega max And a maximum number of iterations k of the algorithm;
2) Initializing a particle swarm; randomly generating initial values of the position and the speed of the particle, wherein the position of the particle represents the position of a node, namely the position of distributed power access, and the speed represents the active power and the reactive power of the corresponding node;
3) Carrying out load flow calculation, and calculating the fitness function value of each particle, wherein the fitness function value is the satisfaction value after multiple targets are converted into a single target;
4) Updating the optimal positions of individuals and populations; each particle is associated with its own p best Making a comparison if p is best When it is good, p is substituted best Else p best Keeping the original shape; the adaptive value p in the current population is calculated best And g best In comparison, if it is better than g best Then it is used to replace g best Otherwise, keep g best Keeping the original shape;
5) Calculating the fitness variance and the variation probability of the current population according to a formula (4-3) and a formula (4-6);
6) Randomly generating r e [0,1 ∈]When r is<P k Carrying out mutation operation according to the formula (4-7), otherwise, turning to the formula (7);
7) Updating the position and the speed of the particles according to the formula (4-1);
8) If the convergence condition is met, stopping searching, and outputting a global optimal solution, otherwise, turning to the step 3);
the distributed power supply optimization configuration solving process based on the AMPSO is shown in the attached figure 2.
Due to the adoption of the technical scheme, compared with the prior art, the method takes the randomness of wind power generation into consideration by considering the benefits of power distribution operators from the planning point of view, and establishes a multi-objective optimization configuration model consisting of the minimum active electric energy loss, the minimum total voltage deviation and the minimum risk cost of the power distribution network; the multi-target normalization is realized by adopting a fuzzy set theory, and the problem that dimensions of all sub-targets are different is solved; for the optimized configuration model established in the text, an Adaptive Mutation Particle Swarm Optimization (AMPSO) is adopted for solving, and due to the introduction of Mutation operation, the situation that a basic Particle Swarm Optimization (PSO) is easy to fall into local Optimization is improved, and finally, an IEEE33 node power distribution system is taken as an example, the optimized configuration model of DG and the selected AMPSO are verified, and the obtained simulation result shows that the model and the algorithm adopted in the text are feasible and effective.
Drawings
FIG. 1 is a schematic representation of a membership function;
FIG. 2 is a schematic diagram of a distributed power optimization configuration solution flow for AMPSO;
FIG. 3 is a schematic diagram of an IEEE33 node power distribution system architecture;
FIG. 4 is a convergence characteristic curve of the fan during the random output optimization;
FIG. 5 is a comparison graph of node voltages before and after fan optimization;
FIG. 6 is a schematic diagram of a convergence curve when optimized using a PSO algorithm;
FIG. 7 is a graph showing the convergence characteristics when optimized using the AMPSO algorithm.
Detailed Description
The invention is described in further detail below with reference to the drawings, but the invention is not limited thereto.
The embodiment of the invention comprises the following steps: in order to verify the effectiveness and feasibility of the proposed optimal configuration model and adaptive variation particle swarm algorithm for the distributed power supply, in this embodiment, an IEEE33 node power distribution system is used as an example, and a programming environment of MatlabR2009a is applied to perform programming calculation on the power flow and the optimization algorithm of the power distribution network, and the specific analysis is as follows:
1. parameter selection
Distribution network parameter
In the embodiment, an IEEE33 node power distribution system is selected as a test object, the influence of accessing distributed power supplies with different capacities at different positions on the power distribution system is tested, and a simulation result is summarized and analyzed, a schematic structural diagram of the IEEE33 node power distribution system is shown in fig. 3, detailed data of branch impedance and load distribution are shown in table 1.1 and table 1.2, the power distribution system has 1 power supply point, namely node 0, 33 nodes in total, 32 branches, the node 0 is a balance node, the rest 32 nodes are load nodes and can be used as the internet access position of power generated by the distributed power supplies, the reference voltage of the head end of the network is 12.66kV, the active load of the whole network before accessing the distributed power supplies is 3715.0kW, and the reactive load is 2300.0kvar.
TABLE 1.1 line parameter Table
TABLE 1.2 load Power Meter
Continuation table 1.2 load power meter
Then the capacity constraint limits for distributed power access: the maximum capacity of the distributed power supply to be connected into the power distribution system should be less than 40% of the total load of the system, so that the maximum active power of the DG connected into the power grid is as follows:
P max ＝P total ×40％＝1486kW
(5-1)
(II) algorithm parameters
In the particle swarm optimization, the population size m =50, the particle dimension n =32, the maximum iteration number is K =300, and the acceleration factor c 1 ＝c 2 =2, inertia weight max and min 0.9 and 0.4, tide, respectivelyThe fans are regarded as PQ nodes with constant power factors in flow calculation, namely DGs connected into the system are regarded as negative loads in the flow calculation, and the power factor is 0.85.
(III) parameters of distributed power supply
The domestic FD20-200kW type wind driven generator adopted in the calculation example has the rated power of 200kW, the cut-in wind speed Vci of 3m/s, the rated wind speed Vr of 13.8m/s and the cut-out wind speed Vco of 25m/s.
2. Analysis of simulation results
The active power loss of the IEEE33 node power distribution system obtained by carrying out load flow calculation through matlab is 175.6542kW/h, the total voltage deviation is 0.07673, the lowest voltage point of the voltage is at a 17 node, and the lowest voltage is 0.9219p.u.
(I) simulation analysis of single random force
Supposing that a certain wind power plant is provided with 8 FD20-200kW wind power generators, the wind power plant is provided with four outgoing lines connected with a power distribution network, no matter how much electric energy is generated by a fan, the wind power plant is not limited by the constraint condition of the distributed power supply access electric energy capacity, AMPSO is adopted to perform one-time optimization solution, matlab programming is utilized, under the local wind field environment, through calculation, the wind speed of the wind field obeys the Weibull distribution shape parameter k =2 and the scale parameter c =15.5, a fan output simulation method is adopted to perform one-time random simulation on the output of 8 fans, and the obtained results are respectively: 200kW, 184.55kW, 200kW, 182kW,175.76kW,148kW,200kW, for a total of 1490.31kW. Then, when a multi-objective optimization configuration model is constructed, because only one time of optimization is carried out, the output of the fan is completely accessed into the system, no risk cost exists, the optimization objective is changed into the minimum active power loss and the minimum total voltage deviation, the optimization result is shown in a table 2.1, the change of membership function corresponding to each sub-objective is shown in a table 2.2, the convergence characteristic curve during algorithm optimization is shown in an attached figure 4, and the comparison result of the node voltage before and after optimization is shown in an attached figure 5.
TABLE 2.1 stochastic output optimization results
Note: the number of the node accessed by the distributed power supply is arranged before the bracket of the planning result item in the table, the number in the bracket is the electric energy capacity accessed to the node, and the unit is as follows: kW.
TABLE 2.2 subgoal membership function value variation
As can be seen from Table 2.1, after the distributed power supply is accessed, the active power loss and the total voltage deviation of the power distribution system are obviously reduced, the active power loss is reduced from 175.6542kW/h to 50.8484kW/h, and the total voltage deviation is reduced from 0.07673 to 0.01158; as can be seen from table 2.2 and fig. 4, when iteration reaches 117 th time, the maximum satisfaction M reaches 0.9848, and the maximum satisfaction is increased from 0.9270 to 0.9848, and the membership function of the sub-targets is gradually increased, which indicates that the overall performance is improved by using the maximum satisfaction method, and each sub-target obtains a better optimization result; as can be seen from the comparison graph of the voltages before and after the optimization in fig. 5, the voltage level of the node after the optimization is significantly increased and is closer to the reference value, the voltage value of the node 17 is increased from 0.9219p.u to 0.979797p.u, which is increased by 6.26%, the lowest node voltage appears at the node 32 and is 0.9636p.u, which is also increased compared with the value at the lowest node voltage before the optimization, so that the overall voltage level of the power distribution system is improved, and the economy is improved.
(II) simulation analysis of planned capacity
From the planning perspective, according to the capacity limit of the distributed power supply to be connected to the power distribution network, it is assumed that a certain wind farm only provides 1200kW of active capacity to the system, that is, the wind farm has 6 FD20-200kW wind power generators, the wind farm has three outgoing lines connected to the power distribution network, and the internet access position and the access capacity of the distributed power supply are determined through the optimal configuration of the distributed power supply. The expected output in the objective function risk cost is then 200kW, if the fan output does not meet 200kW, it will be powered by the superior power supply.
1) Comparison of PSO with AMPSO
Under the conditions, planning for the past ten years is carried out, the PSO method and the AMPSO method are respectively adopted to solve the previously established distributed power supply optimization configuration model, the optimization result only outputs active power loss and total voltage deviation, and the optimization result is shown in a table 2.3. Corresponding to the data in table 2.3, the data of the voltages at the nodes after optimization by PSO and AMPSO are shown in table 2.4, and the convergence characteristic curves when the corresponding PSO and AMPSO are optimized are shown in fig. 6 and 7.
As can be seen from tables 2.3 and 2.4, although the total voltage deviation in the optimization result of the particle swarm algorithm is smaller than the numerical value of the adaptive variation particle swarm algorithm, the difference is small, and in fig. 6, when the particle swarm algorithm iterates 33 times, the maximum satisfaction M reaches 0.9073, and then the algorithm is not changed all the time, falls into the local optimum, and does not reach the global optimum solution, so that the optimization effect is poor; when the AMPSO is adopted for optimization, the value M of the maximum satisfaction degree reaches the optimal 0.9375 after the iteration is carried out for 254 times, the iteration times and the maximum satisfaction degree are better than those of the maximum satisfaction degree adopting the particle swarm algorithm, the variation probability which is changed in a self-adaptive mode according to the variance of the group fitness degree is introduced, when the algorithm is trapped into a local optimal solution (multiple fluctuation can be seen in the AMPSO optimization from a convergence diagram), the speed and the position of particles are changed, the diversity of the group is kept, premature convergence is overcome to a certain extent, a better convergence result is obtained, and the feasibility and the effectiveness of adopting the AMPSO for optimal configuration of DGs are demonstrated.
TABLE 2.3 optimization of PSO and AMPSO
Note: the number of the node accessed by the distributed power supply is arranged before the parenthesis of the planning result item in the table, the number in the parenthesis is the electric energy capacity accessed to the node, and the unit is as follows: kW.
TABLE 2.4 Voltage (p.u) at each node in the system
TABLE 2.4 Voltage (p.u) of each node in the system
2) Impact of Risk cost
Under the conditions given by the second step, planning for a decade is carried out, active power loss and total voltage deviation and active power loss, total voltage deviation and risk cost are respectively taken as objective functions, namely analysis is carried out by taking risk cost into consideration and risk cost into consideration, and optimization solution is carried out by adopting AMPSO, wherein the optimization result of the three objective functions is the optimal result obtained by the AMPSO in the step 1), and the optimization results of two objectives and three objectives are shown in a table 2.5.
TABLE 2.5 optimization results for Dual and triple targets
As can be seen from table 2.5, the optimization with the dual targets can obtain a better optimization effect than the optimization with the three targets, that is, the active power loss and the total voltage deviation of the power distribution system are small, so that the risk cost is illustrated, that is, the randomness of the fan output is considered, so that the benefit of the power distribution operator is reduced, but the distributed power supply truly reflecting the renewable energy is easily affected by the factors such as the natural environment, so that the output has the characteristic of randomness.
Claims (3)
1. An optimal configuration method of a distributed power supply is characterized in that: according to the method, from the planning perspective, by considering the benefits of a power distribution operator and the randomness of the power generation output of the distributed power supply, the optimized configuration of the distributed power supply is researched, namely, the internet access position and the capacity of the distributed power supply are solved, a multi-objective optimized configuration model consisting of the minimum active power loss, the minimum total voltage deviation and the minimum risk cost of the power distribution network is established, and constraint conditions are set and optimized for the multi-objective optimized configuration model; the multi-objective normalization is realized by adopting a fuzzy set theory, the relationship among all sub-objectives is well coordinated, and the effect of overall optimization is further achieved; solving through a self-adaptive variation particle swarm algorithm;
the multi-objective optimization configuration model specifically comprises the following steps:
1. the active electric energy loss is minimum;
the distribution network voltage grade is lower, the R/X value is larger, the network loss is larger in load flow calculation, and the network loss can be reduced by reasonably accessing a distributed power supply, so that the minimum network loss is taken as a target function to bring better economic benefit, and since the minimum network loss is taken as an optimization target from the planning angle, the target function with the minimum active energy loss is as follows:
in the formula: I.C. A i Is the current of branch i; n is the total branch number of the system; r i Is the branch resistance, h is the hours of the programming period;
2. the optimization target model with the minimum total voltage deviation is as follows;
the out-of-limit of the node voltage amplitude value can affect the normal work of a user and the safety of a system, so that the node voltage can be closer to a reference value by taking the total node voltage deviation as a target function, the voltage level of a power distribution network is guaranteed, and the effect of improving the voltage quality is achieved; the objective function for minimum total voltage deviation is:
in the formula: detV is the total voltage deviation; n is the number of network nodes; u shape b Is a reference voltage; u shape i Is the voltage value of the ith node;
3. the risk cost is minimum;
the distributed power supply of renewable energy is easily affected by natural environment or other factors, so that the output of the distributed power supply has the characteristics of randomness, volatility and uncertainty, the randomness of the output of the distributed power supply is reflected through risk cost, when the generated energy of the distributed power supply does not reach an expected value, the cost representation of power supply is carried out through a power grid, namely a superior power supply, and the objective function of the risk cost is as follows:
C risk ＝P·C·t·n
in the formula: p is the probability that the output of the distributed power supply with randomness is smaller than the expected output; c is the cost of power supply by the superior power supply when the output of the distributed power supply is smaller than the expected output; t is the annual operating time of the distributed power supply; and n is the planning age.
2. The method according to claim 1, wherein the method comprises the following steps: and the constraint conditions of the multi-objective optimization configuration model are inequality constraints and comprise node voltage constraints, distributed power supply access constraints and branch current constraints.
3. The method according to claim 1, wherein the method comprises: the method for realizing multi-target normalization by adopting the fuzzy set theory firstly improves the index of difference by carrying out maximum processing on the minimum value of the membership of the multi-targets to achieve the aim of improving the overall performance of the system, the adopted membership function is a piecewise linear function, for the optimization containing a plurality of objective functions, the optimization targets are often contradictory to each other and are difficult to simultaneously achieve the optimization, and under the condition that the importance of each optimization target is the same, the maximum satisfaction degree method is adopted for solving, namely, the larger the minimum value of the membership function of the optimization target is, the larger the overall satisfaction degree is.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510464144.7A CN105071433B (en) | 2015-07-31 | 2015-07-31 | A kind of configuration scheme of distributed generation resource |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510464144.7A CN105071433B (en) | 2015-07-31 | 2015-07-31 | A kind of configuration scheme of distributed generation resource |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105071433A CN105071433A (en) | 2015-11-18 |
| CN105071433B true CN105071433B (en) | 2018-04-20 |
Family
ID=54500746
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510464144.7A Active CN105071433B (en) | 2015-07-31 | 2015-07-31 | A kind of configuration scheme of distributed generation resource |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105071433B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105703397B (en) * | 2016-04-07 | 2018-03-02 | 沈阳农业大学 | A kind of distributed power source Optimal Configuration Method for considering countryside low-voltage electric distribution network low-voltage and administering |
| CN106058858B (en) * | 2016-06-30 | 2018-06-19 | 中国南方电网有限责任公司电网技术研究中心 | A kind of distribution optimization method and device |
| CN106845626B (en) * | 2017-01-05 | 2020-04-17 | 国网福建省电力有限公司 | DG optimal configuration application method based on mixed frog-leaping particle swarm |
| CN107038499B (en) * | 2017-04-07 | 2021-04-30 | 山东大学 | Global energy optimization configuration method based on minimum deviation method |
| CN107067344B (en) * | 2017-05-27 | 2020-12-25 | 东北电力大学 | Distributed power supply multi-target access planning method considering power distribution network harmonic management |
| CN109309392A (en) * | 2017-07-28 | 2019-02-05 | 国网江苏省电力公司常州供电公司 | Distributed power source output power Optimal Configuration Method based on particle swarm algorithm |
| CN110503227A (en) * | 2018-05-18 | 2019-11-26 | 中国电力科学研究院有限公司 | A kind of method for optimizing configuration and device of distributed busbar protection |
| CN111509750B (en) * | 2020-04-27 | 2021-02-19 | 湖南经研电力设计有限公司 | Power grid side energy storage system capacity configuration optimization method |
| CN112630564B (en) * | 2020-12-07 | 2023-02-28 | 国网宁夏电力有限公司电力科学研究院 | Transformer DGA fault diagnosis method based on neighborhood rough set and AMPOS-ELM |
| CN113361805B (en) * | 2021-06-30 | 2023-02-07 | 国网内蒙古东部电力有限公司经济技术研究院 | Power distribution network planning method and system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5527712B2 (en) * | 2010-03-10 | 2014-06-25 | 東京電力株式会社 | Distributed power control method and distributed power control system |
| CN104659816B (en) * | 2015-03-13 | 2018-04-20 | 贵州电力试验研究院 | The excellent method of completing the square of distribution system is accessed based on the distributed generation resource for improving particle cluster algorithm |
-
2015
- 2015-07-31 CN CN201510464144.7A patent/CN105071433B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN105071433A (en) | 2015-11-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105071433B (en) | A kind of configuration scheme of distributed generation resource | |
| Wen et al. | Economic allocation for energy storage system considering wind power distribution | |
| Niknam et al. | Scenario-based multiobjective distribution feeder reconfiguration considering wind power using adaptive modified particle swarm optimisation | |
| Niknam et al. | A practical multi-objective PSO algorithm for optimal operation management of distribution network with regard to fuel cell power plants | |
| Hong et al. | Optimal VAR planning considering intermittent wind power using Markov model and quantum evolutionary algorithm | |
| CN108471143A (en) | Micro-grid multi-energy method for optimizing scheduling based on positive and negative feedback particle cluster algorithm | |
| Tang et al. | Study on day-ahead optimal economic operation of active distribution networks based on Kriging model assisted particle swarm optimization with constraint handling techniques | |
| Sedighizadeh et al. | The imperialist competitive algorithm for optimal multi-objective location and sizing of DSTATCOM in distribution systems considering loads uncertainty | |
| Hong et al. | Chaotic PSO-based VAR control considering renewables using fast probabilistic power flow | |
| CN113378100B (en) | Power distribution network source network load storage collaborative optimization scheduling model and method considering carbon emission | |
| Rao et al. | Optimum allocation and sizing of distributed generations based on clonal selection algorithm for loss reduction and technical benefit of energy savings | |
| CN112561273B (en) | Active power distribution network renewable DG planning method based on improved PSO | |
| Keerio et al. | Multi-objective optimal reactive power dispatch considering probabilistic load demand along with wind and solar power integration | |
| Ma et al. | Optimal economic operation of islanded microgrid by using a modified pso algorithm | |
| Li et al. | Active distribution network operational optimization problem: A multi-objective tuna swarm optimization model | |
| CN110135640B (en) | Wind power distribution network optimal scheduling method based on fuzzy clustering improved harmony algorithm | |
| Liu et al. | Multi-objective optimal allocation of TCSC for power systems with wind power considering load randomness | |
| Khosravi et al. | Distribution of optimum reactive power in the presence of wind power plant and considering voltage stability margin using genetic algorithm and Monte Carlo methods | |
| Wang et al. | Minimum load-shedding calculation approach considering loads difference | |
| Kaviani-Arani | Optimal placement and sizing of distributed generation units using co-evolutionary particle swarm optimization algorithms | |
| CN113919721A (en) | Distributed power supply multi-target planning method with coordinated reliability and economy | |
| Jiao et al. | An enhanced quantum-behaved particle swarm algorithm for reactive power optimization considering distributed generation penetration | |
| Estabragh et al. | A novel approach for optimal allocation of distributed generations based on static voltage stability margin | |
| Fathi et al. | Multi-Objective optimization of islanded microgrids | |
| Yasin et al. | Multiobjective quantum-inspired evolutionary programming for optimal location and sizing of distributed generation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |