CN110601179A - Receiving-end power grid consumption optimization method for wind power participating in frequency modulation - Google Patents

Abstract

The invention discloses a receiving-end power grid consumption optimization method in which wind power participates in frequency modulation. The method is as follows: firstly construct a receiving-end power grid consumption calculation model, set system information and initial parameters, and randomly generate control variable initial values and particle initial speeds; According to the dynamic power flow, the frequency value of the wind turbine participating in the frequency modulation is calculated after a certain disturbance and after the lag time; then the system flow of the particles under each wind speed sample is calculated to determine whether the constraint is satisfied; the constraint is added to the fitness function to calculate the fitness of each particle Degree value, obtain the individual optimal value and the global optimal value, and update the velocity and position of each particle; then judge whether the maximum number of iterations is satisfied, if not, return to recalculate the system power flow of the particle under each wind speed sample, if satisfied, then Output the optimal decision variables and the limit value of the receiving capacity of the power grid. The invention can effectively calculate the absorptive capacity of the receiving end power grid under the condition that wind power participates in frequency regulation.

Description

Optimal method for receiving power grid consumption with wind power participating in frequency regulation

technical field

The invention belongs to the field of electric power systems and automation thereof, in particular to a receiving-end power grid consumption optimization method in which wind power participates in frequency modulation.

Background technique

At present, my country's wind power generation is in a stage of rapid development. Most of the 10-million-kilowatt-scale wind power bases that have been built or are under construction are mostly far away from the central network, often at the edge or end of the power grid, and the corresponding grid structure is relatively weak. And the power structure Relatively simplistic, the large-scale access of wind power will have a great impact on the safe and stable operation of the regional power grid. At the same time, the fluctuation of wind will make the output power of wind farms fluctuate. The proportion of the grid is getting higher and higher, and its influence on the power grid is also gradually expanding from the local area. Therefore, it is of great significance to further strengthen the construction and planning of the power grid and study the consumption capacity of the receiving power grid for the development and utilization of wind power in my country.

Conventional water-fired power units can adjust the output of the unit through the governor after the grid frequency changes. Wind turbines cannot respond to changes in grid frequency, and will inevitably lead to a decline in the primary frequency regulation capability of the grid after replacing some conventional wind turbines. It can be seen that as the proportion of wind power in the power grid increases, the above-mentioned problems that endanger the safe operation of the power grid will inevitably be exacerbated. Wind power participation in grid frequency regulation is an important method to ensure better grid-connected operation of wind power and improve the level of wind power consumption.

However, in the traditional accommodation problem, wind power participates in power flow calculation as a disturbance source, and after wind power participates in frequency regulation, power flow analysis at different time sections is required. Existing consumption optimization performs multiple dynamic simulation experiments with assumed wind power injection power for frequency, voltage, peak shaving capacity, static security and transient stability and other constraints on wind power access, and constantly corrects the assumed values. Until the constraints are met, the wind power penetration power limit is determined. The disadvantages of this method are: 1) The amount of calculation is very large, and the wind power injection power needs to be manually adjusted, and there is currently no adjustment principle to follow, and the adjustment based on experience is highly subjective, which is not conducive to operation, and the engineering practicability is poor ; 2) Generally only for a specific constraint factor, it is impossible to comprehensively consider multiple constraints affecting wind power access; 3) It is impossible to fully consider various operating modes of the system and wind speed conditions of the wind farm.

Contents of the invention

The purpose of the present invention is to provide a receiving-end power grid consumption optimization method in which wind power participates in frequency regulation.

The technical solution to realize the object of the present invention is: a receiving end power grid consumption optimization method in which wind power participates in frequency regulation, comprising the following steps:

Step 1. Construct the consumption calculation model of the receiving power grid considering the participation of wind power in frequency regulation, and set the system information and the initial parameters of the particle swarm optimization algorithm;

Step 2, randomly generating the initial value of the control variable and the initial velocity of the particle;

Step 3. According to the dynamic power flow, calculate the frequency value of the wind turbine participating in the frequency modulation under a certain disturbance and after the lag time;

Step 4. Calculate the system power flow of the particles under each wind speed sample, and judge whether the constraints are satisfied;

Step 5. Add the constraints to the fitness function, calculate the fitness value of each particle, obtain the individual optimal value and the global optimal value, and update the velocity and position of each particle;

Step 6. Determine whether the maximum number of iterations is satisfied, if not, return to step 4; if satisfied, proceed to step 7;

Step 7. Output the optimal decision variable and the limit value of the receiving capacity of the power grid at the receiving end.

Compared with the prior art, the present invention has significant advantages in that: (1) the present invention can effectively calculate the limit value of the receiving capacity of the receiving power grid when wind power participates in frequency regulation, and has guiding significance for the planning and design of wind farms; (2) In view of the characteristics of wind power participating in frequency modulation, the traditional optimal power flow cannot be directly applied to the calculation of wind power access capacity. In the case of wind power participating in frequency regulation, the limit value of the receiving capacity of the receiving power grid is calculated accurately.

Description of drawings

Fig. 1 is a flow chart of the receiving-end power grid consumption optimization method in which wind power participates in frequency regulation according to the present invention.

Fig. 2 is a frequency change diagram of wind power participating in frequency modulation in the method of the present invention.

Detailed ways

The receiving-end power grid consumption optimization method for wind power participating in frequency modulation of the present invention considers the characteristics of wind power participating in frequency modulation, and solves the problem that the traditional optimal power flow cannot be directly applied to the calculation of wind power access capacity. Firstly, a consumption calculation model of the receiving power grid considering the participation of wind power in frequency regulation is constructed, and a constrained programming model is established. Finally, the improved particle swarm optimization algorithm is used to solve the overall model, and the optimal decision variables and limit values of consumption capacity are output.

With reference to Fig. 1, the receiving end power grid consumption optimization method for wind power participating in frequency regulation proposed by the present invention, the specific steps are as follows:

Step 1. Construct the consumption calculation model of the receiving power grid considering the participation of wind power in frequency regulation, and set the system information and the initial parameters of the particle swarm optimization algorithm, as follows:

The consumption optimization model of the receiving end power grid in which wind power participates in frequency regulation, under the premise of satisfying the system power flow equation constraints and a series of inequality constraints for reliable and safe operation of the system, takes the maximum sum of the installed capacity of each wind farm that the system can accommodate as the goal, and selects The number of fans equipped in each wind farm and the active power of conventional units are used as decision variables for optimal adjustment. Essentially, this problem is a multivariate, multiconstraint, nonlinear mixed integer programming problem with random variables.

The model aims at maximizing the sum of the installed capacity of each wind farm that can be accommodated by the system, and the objective function is:

In the formula, m is the number of wind farms, n _i is the number of wind turbines in the i-th wind farm, and P _NWi is the rated power of the wind turbines in the i-th wind farm.

The equality constraints are the power flow equations of the system:

In the formula, P _Gi , Q _Gi are the active and reactive power of the conventional generating set at node i respectively, P _Wi , Q _Wi are the active and reactive power of the wind farm at node i respectively, P _Li , Q _Li are respectively is the active and reactive load at node i, U _i , U _j , θ _ij are the voltage amplitude and phase angle difference between node i and node j respectively, G _ij , B _ij are the real part of the system admittance matrix and imaginary part, C _PQ , C _PV are respectively the set of PQ, PV nodes;

Inequality constraints include constraints on decision variables and state variables, where the decision variables are the number of fans and the power of conventional units, and the constraints are:

In the formula, is the maximum number of wind turbines in the i-th wind farm, are the minimum and maximum active power of the i-th conventional generator set, respectively, and C _G is the set of conventional generator sets.

The state variable is the dependent variable of the decision variable, including system frequency, node voltage amplitude, reactive power of conventional generator sets, line power flow, upper and lower spinning reserves of the system, and rampability constraints of conventional generator sets. The state variable constraints are:

In the formula, f ^min and f ^max are the minimum and maximum frequencies of the system respectively; are the minimum and maximum voltage amplitudes of node i, respectively; are the minimum and maximum reactive power of the i-th conventional generator set; P _Li is the power flow on the i-th line, is the maximum power flow limit on the i-th line; P _Gi is the active power of the i-th conventional generator set, is the minimum active power of the i-th conventional generating set; are the upper and lower spinning reserves of the system, generally 5% of the total system load; r _Gi is the maximum climbing capacity of the conventional unit i considering the upper and lower limits of its power, is the power sample sequence of wind farm j.

Set the system node parameters and power flow parameters, as well as the initial parameters of the particle swarm algorithm;

Step 2, randomly generating the initial value of the control variable and the initial velocity of the particle;

Randomly generate N D-dimensional particles within the scope of the control variable, and the initial velocity of each particle. For the integer control variable, that is, the maximum number of fans in the wind farm, the initial value of the particle needs to be rounded.

Step 3. According to the dynamic power flow, calculate the frequency value of the wind turbine participating in the frequency modulation after a certain disturbance (load increase or decrease by 10% or 20%) and after the lag time (here is set to 5s), the details are as follows:

Establish the primary frequency modulation characteristics of variable speed wind turbines using the coordinated frequency modulation control strategy of overspeed and pitch change. Through the unbalanced power segment distribution method, the Euler method is used to solve the system frequency differential equation, and the system is calculated under a certain disturbance. After the delay time, the system The frequency value is simulated as shown in Figure 2. The frequency curve drops before the set delay time of 5s, and the power correction is performed after 5s, and the frequency starts to rise.

Step 4. Calculate the system power flow of particles under each wind speed sample, and judge whether the equality constraints and inequality constraints in step 1 are satisfied, as follows:

After obtaining the sample matrix of wind speed, the following piecewise function can be used to simplify the expression of the relationship between the active power of the wind turbine and the wind speed:

In the formula, v _in , v _out , v _N are the cut-in wind speed, cut-out wind speed and rated wind speed of the wind turbine, respectively, and p, p _N are the actual output power and rated output power of the wind turbine, respectively.

Step 5. Add constraints to the fitness function, calculate the fitness value of each particle, obtain the individual optimal value and the global optimal value, and update the speed and position of each particle, as follows:

Calculate whether the wind speed sample of each initial particle satisfies the constraint conditions, and calculate the fitness function as:

In the formula, when x=1, it means that the particle satisfies the constraint condition, and x=0 is the opposite, so only when the particle satisfies the constraint condition, the fitness function can reach the maximum value.

Assuming that the particle swarm is composed of N particles, and each particle is defined as a D-dimensional space, the velocity and position of particle i can be updated according to the following formula:

In the formula, i=1, 2..., M is the number of particles; d=1, 2..., D is the dimension of the particle, that is, the dimension of the solution of the problem to be optimized; c ₁ , c ₂ is the learning factor; r ₁ and r ₂ are random numbers uniformly distributed on (0,1); ω is the inertia weight; Respectively, the velocity and position of particle i at the kth iteration; are the individual historical optimal value of particle i and the global historical optimal value of all particles, respectively.

Among them, the inertia weight ω adopts a nonlinear decreasing strategy and decreases with a concave function:

ω＝(ω _start -ω _end )(t/t _max ) ² +(ω _end -ω _start )(2t/t _max )+ω _start (8)

In the formula, ω _start and ω _end are the initial inertia weight and end inertia weight respectively; t and t _max are the current iteration number and the maximum iteration number respectively.

For the integer decision variable n _i , in order to ensure that the speed and position of the k+1th iteration are also integers, the formula for updating the speed is:

In the formula, int represents the rounding function; Represents a random number uniformly distributed on the interval, when hour, Take the interval uniformly distributed random numbers on Take the interval A random number uniformly distributed on ; when hour, Take the interval uniformly distributed random numbers on Take the interval uniformly distributed random numbers on

After each speed update, judge whether the speed exceeds the limit, if it exceeds the limit, correct the speed according to the following formula:

For a particle whose search space is limited to [X _min ,X _max ], its maximum velocity v _max is:

v _max = λ(X _max -X _min )/2, 0.1≤λ≤1 (11)

Step 6. Determine whether the maximum number of iterations is satisfied, if not, return to step 4; if satisfied, proceed to step 7;

Step 7. Output the optimal decision variable number of wind turbines and the power of conventional units, and the limit value H of the receiving capacity of the power grid at the receiving end.

The receiving capacity of the receiving grid is defined as the percentage of the maximum installed capacity of wind farms that the system can accept in the system load. According to the output decision variables, the limit value of the receiving capacity of the power grid at the receiving end is calculated.

The receiving-end power grid consumption optimization method for wind power participating in frequency modulation proposed by the present invention combines constraint programming and particle swarm algorithm and has a good application prospect in power system simulation; after adding wind power to participate in frequency modulation, with the help of its special control characteristics , can effectively improve the shortcomings that the model cannot be applied to practice; the invention can effectively calculate the limit value of wind power penetration power of each access point when wind power participates in frequency modulation, and has certain guiding significance for the planning and design of wind farms.

Claims (4)

1. A receiving-end power grid absorption optimization method for wind power participating in frequency modulation is characterized by comprising the following steps:

step 1, constructing a receiving-end power grid absorption calculation model considering wind power participation frequency modulation, and setting system information and particle swarm algorithm initial parameters;

step 2, randomly generating a control variable initial value and a particle initial speed;

step 3, calculating a frequency value of the wind turbine generator participating in frequency modulation under certain disturbance and after the lag time according to the dynamic power flow;

step 4, calculating the system load flow of the particles under each wind speed sample, and judging whether the constraint is met;

step 5, adding the constraint into a fitness function, calculating the fitness value of each particle, acquiring an individual optimal value and a global optimal value, and updating the speed and the position of each particle;

step 6, judging whether the maximum iteration times are met, and if not, returning to the step 4; if yes, performing step 7;

and 7, outputting the optimal decision variables and the receiving end power grid absorption capacity limit value.

2. The receiving-end power grid absorption optimization method for wind power participation frequency modulation according to claim 1, wherein the receiving-end power grid absorption calculation model in the step 1 takes maximization of sum of installed capacities of all wind power plants which can be received by a system as a target, and the target function is as follows:

wherein m is the number of wind power plants, n_iIs the number of fans in the ith wind farm, P_NWiRated power of a fan in the ith wind power plant;

the equality constraint is the power flow equation of the system:

in the formula, P_Gi、Q_GiActive and reactive power, P, respectively, of a conventional generator set at node i_Wi、Q_WiRespectively the active and reactive power, P, of the wind farm at node i_Li、Q_LiRespectively active and reactive loads, U, at node i_i、U_j、θ_ijVoltage amplitude and phase angle difference, G, of node i and node j, respectively_ij、B_ijRespectively the real and imaginary parts, C, of the system admittance matrix_PQ、C_PVSets of PQ, PV nodes, respectively;

the inequality constraints comprise constraints of decision variables and state variables, wherein the decision variables are the number of fans and the power of a conventional unit, and the constraints are as follows:

in the formula (I), the compound is shown in the specification,the maximum number of the fans in the ith wind power plant,minimum and maximum active power, C, of the ith conventional generator set, respectively_GIs a set of conventional generator sets;

the state variables are dependent variables of the decision variables, and include system frequency, node voltage amplitude, reactive power of a conventional generator set, line power flow, up-down rotation standby of the system and climbing capacity constraint of the conventional generator set, and the state variable constraint is as follows:

in the formula (f)^min、f^maxRespectively, the minimum and maximum frequencies of the system;minimum and maximum voltage amplitudes for node i, respectively;respectively the minimum and maximum reactive power of the ith conventional generator set; p_LiFor the power flow on the ith line,the maximum limit value of the power flow on the ith line is set; p_GiThe active power of the ith conventional generator set,the minimum active power of the ith conventional generator set;respectively rotating the upper part and the lower part of the system for standby; r is_GiThe maximum climbing capacity of the conventional unit i after the upper and lower power limits are restrained is considered,is a sequence of power samples for wind farm j.

3. The receiving-end power grid digestion optimization method for wind power participation frequency modulation according to claim 1, wherein the step 2 specifically comprises: n D-dimensional particles are randomly generated in the control variable range, the initial speed of each particle is generated, and the initial value of each particle needs to be rounded for an integer control variable, namely the maximum configuration number of fans in the wind power plant.

4. The receiving-end power grid digestion optimization method for wind power participation frequency modulation according to claim 1, wherein the step 5 specifically comprises:

calculating whether the wind speed sample of each initial particle meets the constraint condition, and calculating a fitness function as follows:

in the formula, when x is 1, the particle satisfies the constraint condition, and when x is 0, the opposite is true, so that the fitness function can reach the maximum value only when the particle satisfies the constraint condition;

assuming that the particle group consists of N particles, each defined as a D-dimensional space, the velocity and position of particle i can be updated according to the following equation:

wherein, i is 1,2, and M is the number of particles; d is the dimension of the particle, i.e. the dimension of the solution of the problem to be optimized; c. C₁、c₂Is a learning factor; r is₁、r₂Random numbers uniformly distributed on (0, 1); omega is the inertial weight;respectively the speed and position of the particle i at the kth iteration;respectively obtaining an individual historical optimal value of the particle i and a global historical optimal value of all the particles;

the inertia weight omega adopts a nonlinear decreasing strategy, and decreases by a concave function:

ω＝(ω_start-ω_end)(t/t_max)²+(ω_end-ω_start)(2t/t_max)+ω_start (7)

in the formula, ω_start、ω_endAre respectively an initial inertial weight and a termination inertial weight; t, t_maxRespectively the current iteration times and the maximum iteration times;

for integer decision variables n_iIn order to ensure that the speed and position in the (k + 1) th iteration are also integers, the formula for updating the speed is as follows:

in the formula, int represents an integer function;random numbers uniformly distributed over the interval are represented whenWhen the temperature of the water is higher than the set temperature,taking intervalsThe random numbers are uniformly distributed on the random number,taking intervalsRandom numbers uniformly distributed thereon; when in useWhen the temperature of the water is higher than the set temperature,taking intervalsThe random numbers are uniformly distributed on the random number,taking intervalsRandom numbers uniformly distributed thereon;

after each speed updating, judging whether the speed exceeds the limit, and if so, correcting the speed according to the following formula:

for search space constraints of [ X ]_min,X_max]Of particles of (2) having a maximum velocity v_maxComprises the following steps:

v_max＝λ(X_max-X_min)/2,0.1≤λ≤1 (10)。