Converter station distributed node reactive compensation optimization method and system based on improved IVY algorithm

Abstract

The invention discloses a reactive compensation optimization method and system for a converter station distribution node based on an improved IVY algorithm, and the method comprises the steps: constructing a load flow calculation model of the converter station distribution node, including a converter station converter model and a network equation of a converter station, and solving a DC system variable of the converter station distribution node; taking the minimum network loss as a target, introducing a penalty function to limit border crossing, constructing a target function, and implementing node power equation constraint to obtain a reactive compensation configuration model of the distributed nodes of the converter station; an IVY algorithm is improved to optimize the reactive compensation configuration model, and an optimal reactive compensation configuration scheme under the current power flow is obtained; evaluating the reactive compensation configuration scheme, and verifying whether the reactive compensation configuration scheme meets all constraint conditions; judging whether the maximum number of iterations is met or not, and if yes, ending the step; the power flow change is detected in real time, and the reactive compensation scheme is obtained again when the power flow changes. According to the method, the compensation effect can be improved, global optimization is realized, the model complexity is reduced, constraint conditions are comprehensively processed, and the algorithm adaptability is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of reactive power compensation of electric power systems, and more specifically, relates to a method and system for optimizing reactive power compensation of distributed nodes of converter stations using an improved IVY algorithm. Background Art

[0002] With the rapid development of the electric power industry and the continuous expansion of the power grid, the DC transmission system has been widely used due to its advantages of long-distance and large-capacity transmission. However, in the process of DC transmission, due to the operating characteristics of equipment such as converters, a large amount of reactive power demand will be generated. If these reactive powers cannot be effectively compensated, it will lead to problems such as system voltage fluctuations, increased line losses, and equipment overload, thereby affecting the stability and reliability of the transmission system. The reactive compensation configuration technology of high-voltage DC converter stations is one of the key technologies in the DC transmission system. Reasonable reactive compensation configuration can effectively reduce the reactive loss of the system and improve the stability and reliability of the transmission system.

[0003] During operation, the DC converter station will generate a large amount of reactive power demand. If the reactive power compensation is insufficient or excessive, it will affect the reactive power balance of the system, which may lead to voltage fluctuations and system instability. Insufficient or excessive reactive power compensation may affect the stability and reliability of the system, and may even cause the system to crash. Therefore, it is necessary to strengthen the monitoring and maintenance of reactive power compensation equipment to ensure that it is always in good operating condition. At the same time, emergency plans and fault handling mechanisms should be formulated to deal with possible emergencies. Through scientific and reasonable reactive power compensation configuration, the reactive power regulation capability and anti-disturbance capability of the system can be improved, thereby ensuring the stable operation of the system.

[0004] The IVY (ivy) algorithm is a new swarm intelligence optimization algorithm, inspired by the orderly coordinated growth and diffusion evolution process of ivy, and mainly simulates the different life stages of ivy, including growth, rise and spread in the ivy plant community. The IVY algorithm shows fast convergence speed and higher solution accuracy when solving optimization problems. This is due to the simulated ivy growth mechanism in the algorithm, which enables the algorithm to quickly locate the optimal solution area in the search space and quickly locate near the maximum power point. The IVY algorithm has better optimization capabilities than genetic algorithms, particle swarm algorithms, gray wolf optimization algorithms, sparrow search algorithms, etc. in terms of convergence speed, solution accuracy and stability.

[0005] In the prior art, there is research on reactive power compensation point distribution and compensation capacity schemes in the process of transient voltage recovery after the AC system is disturbed or faulty in an AC / DC hybrid system, in order to help the system transient voltage recovery. The AC / DC transmission system is divided into two parts, DC and AC, for reactive power point screening. The load rate of the AC system, the transient voltage recovery time, the interaction between DC converter stations, and the role of the generator as a reactive power source are mainly considered. The network contraction model and performance indicators are used to design reactive power compensation point distribution and compensation capacity optimization.

[0006] The existing technology mainly has the following disadvantages:

[0007] (1) Limited compensation effect: Traditional reactive power compensation methods are often based on fixed compensation strategies or simple feedback control, which makes it difficult to make precise adjustments based on the real-time operating status of the system. This results in less than ideal compensation effects and is unable to fully meet the system’s demand for reactive power.

[0008] (2) Lack of global optimization: Most existing methods only focus on the compensation of local nodes and lack optimization consideration of the global reactive power distribution of the entire DC converter station system, which may lead to over-compensation or under-compensation of some nodes, affecting the overall operating efficiency of the system.

[0009] (3) High model complexity: The DC converter station system involves multiple variables and complex physical processes. Existing models often require large amounts of computation and take a long time to process these complex relationships, making them difficult to apply in real time.

[0010] (4) Inadequate processing of constraints: When configuring reactive power compensation, multiple constraints need to be considered. Existing methods are often not comprehensive or accurate enough in processing these constraints, which may result in configuration results that do not meet actual operating requirements.

[0011] (5) Poor algorithm adaptability: The existing algorithms have poor adaptability and robustness when dealing with DC converter station systems of different scales and operating conditions, and it is difficult to ensure that satisfactory compensation configuration results can be obtained under various circumstances. Summary of the invention

[0012] In order to solve the deficiencies in the prior art, the present invention provides a method and system for optimizing reactive power compensation of distributed nodes in a converter station using an improved IVY algorithm, so as to improve the operating efficiency and stability of a DC converter station system.

[0013] The present invention adopts the following technical solution.

[0014] A first aspect of the present invention provides a method for optimizing reactive power compensation of distributed nodes in a converter station using an improved IVY algorithm, comprising the following steps:

[0015] Step 1, constructing a power flow calculation model of the converter station distribution nodes, including a converter station converter model and a network equation of the converter station;

[0016] Step 2, solving the power flow calculation model to obtain the DC system variables of the converter station distribution nodes;

[0017] Step 3: According to the DC system variables, the reactive power compensation configuration of the distributed nodes is optimized, the minimum network loss of the converter station is optimized, and the penalty function is introduced to limit the constraint variable from crossing the boundary, the objective function is constructed, and the node power equation constraint is implemented on the objective function to obtain the reactive power compensation configuration model of the distributed nodes of the converter station;

[0018] Step 4, using the improved IVY algorithm to optimize the reactive power compensation configuration model of the distributed nodes of the DC converter station, and obtain the optimal reactive power compensation configuration scheme under the current power flow;

[0019] Step 5: Evaluate the optimal reactive power compensation configuration scheme to verify whether it meets all constraints;

[0020] Step 6, determine whether the maximum number of iterations is met, if so, end the step; detect the flow change in real time, and return to step 3 when the flow changes.

[0021] Preferably, in step 1, the converter model of the converter station is expressed by the following formula:

[0022]

[0023] Where:

[0024] I b is the converter current of the DC converter station;

[0025] It refers to the distributed node current;

[0026] U b Indicates the converter voltage;

[0027] It refers to the voltage of the distribution node;

[0028] D r It refers to the transformer ratio of the converter;

[0029] Distribution node control angle;

[0030] λ b is the converter control angle;

[0031] is the transformer ratio of the distribution node;

[0032] Refers to the distributed node power;

[0033] The network equation of the converter station is expressed as follows:

[0034] B(U w ,A)=0

[0035] Where:

[0036] B represents the converter station control coefficient;

[0037] U w Represents the bus voltage of the DC converter station;

[0038] A represents the DC system variable.

[0039] Preferably, in step 3, the penalty functions include: a DC current out-of-bounds penalty function, a converter reactive output out-of-bounds penalty function, a distribution node voltage out-of-bounds penalty function, a converter temperature out-of-bounds penalty function and a line thermal capacity out-of-bounds penalty function.

[0040] Preferably, the constraint variables include distribution node voltage, converter reactive output, DC converter current, converter temperature and line thermal capacity;

[0041] The DC current out-of-bounds penalty function is expressed as follows:

[0042]

[0043] Where:

[0044] ε1 refers to the out-of-bounds penalty coefficient corresponding to the DC current;

[0045] H x is the number of DC network nodes;

[0046] I m It refers to the current passing through the DC network node of the converter;

[0047] I max I m The maximum value of

[0048] I min I m The minimum value of

[0049] The inverter reactive output out-of-bounds penalty function is expressed as follows:

[0050]

[0051] Where:

[0052] ε2 is the corresponding out-of-bounds penalty coefficient of the converter reactive output;

[0053] Hy It refers to the total number of inverters;

[0054] E j It refers to the reactive output of the converter;

[0055] E jmax Represents E j Maximum value;

[0056] E jmin Refers to E j Minimum value;

[0057] The penalty function for voltage out-of-bounds distribution node is expressed as follows:

[0058]

[0059] Where:

[0060] ε3 refers to the corresponding out-of-bounds penalty coefficient of the distribution node voltage;

[0061] H j Refers to the total number of distribution nodes;

[0062] U i It refers to the voltage of the distribution node;

[0063] U imax Refers to the maximum value of the voltage at the distribution node;

[0064] U imin Refers to the minimum value of the voltage at the distribution node;

[0065] The inverter temperature out-of-bounds penalty function is expressed as follows:

[0066]

[0067] Where:

[0068] ε4 is the corresponding out-of-bounds penalty coefficient of the converter temperature;

[0069] T j is the temperature of converter j;

[0070] T max,j is its maximum allowable temperature;

[0071] The line thermal capacity out-of-bounds penalty function is expressed as follows:

[0072]

[0073] Where:

[0074] ε5 is the corresponding out-of-bounds penalty coefficient of the line thermal capacity;

[0075] Iij is the actual current of line ij;

[0076] I thernal,ij is the thermal capacity current of line ij;

[0077] Number of lines H x is the number of DC network nodes. Preferably, in step 3, the objective function is expressed as follows:

[0078] minX=f1+f2+f3+f4+f5+Q1+Q2

[0079] Where:

[0080] minX refers to the minimum network loss of the converter station;

[0081] Q1 refers to the active power loss of the distributed nodes of the DC converter station;

[0082] Q2 refers to the reactive power loss of the DC converter station converter.

[0083] Preferably, in step 3, the node power equation constraint is expressed as follows:

[0084]

[0085] Where:

[0086] P i1 is the corresponding active power output of the converter node;

[0087] P i2 is the corresponding active load power of the distribution node;

[0088] P i3 is the actual active input of the DC network node;

[0089] U m is the converter voltage;

[0090] H j is the total number of distribution nodes;

[0091] U n is the average voltage of the distribution node;

[0092] W mn It refers to the conductance phase angle difference between the converter node and the distribution node;

[0093] is the voltage phase angle difference between the distribution node and the converter node;

[0094] S mn is the susceptance phase angle difference between the distribution node and the converter node;

[0095] Q i1 is the actual reactive power output of the converter node;

[0096] Q j1 is the actual reactive power compensation capacity of the distribution node;

[0097] Q i2 is the actual reactive load power of the distribution node;

[0098] Q i3 is the actual reactive power input of the DC network node;

[0099] H y is the total number of converters.

[0100] Preferably, in step 4, optimizing the reactive power compensation configuration model of the distributed nodes of the DC converter station by using the improved IVY algorithm further includes:

[0101] The initial position of the IVY population in the search space is determined by the following formula:

[0102] I i =I min +rand(1,D)⊙(I max -I min ),i=1,…,Npop

[0103] Where:

[0104] rand(1,D) represents a vector of dimension D of uniformly distributed random numbers in the interval [0,1];

[0105] I max and I min are the upper and lower bounds of the search space respectively;

[0106] ⊙ represents the element-wise product;

[0107] Npop is the population size.

[0108] Preferably, in step 4, optimizing the reactive power compensation configuration model of the distributed nodes of the DC converter station by using the improved IVY algorithm further includes:

[0109] The growth rate is calculated according to the following difference equation,

[0110]

[0111] Where:

[0112] Vector ΔGv i (t) and ΔGvi(t+1) represent the growth rates of the discrete-time system at time t and time t+1;

[0113] rand is a random real number in the interval [0,1];

[0114] rand 2 is a random number of a random variable whose probability density function is equal to

[0115] N(1,D) represents a random vector with dimension D, which is a random number in the standard normal distribution;

[0116] is the mean of the objective function values ​​of all individuals in the population;

[0117] is a distribution vector, each element of which is the mean of all individuals corresponding to the same constituent element in the population;

[0118] For point division, is the element-wise division of vector u by vector v.

[0119] Preferably, in step 4, optimizing the reactive power compensation configuration model of the distributed nodes of the DC converter station by using the improved IVY algorithm further includes:

[0120] Perform survivor selection, when individual I i The objective function value is less than the current optimal individual I best When the objective function value is β times, β = (2 + rand) / 2, rand is a random real number in the interval [0,1]. The width of ivy branches and leaves is expressed as follows:

[0121]

[0122] Where:

[0123] Update results for the current individual;

[0124] I ii For individual I i Neighboring individuals;

[0125] N(1,D) represents a random vector of dimension D, whose components are random numbers from the standard normal distribution;

[0126] |N(1,D)| is a vector of dimension D consisting of the absolute values ​​of the components of N(1,D);

[0127] ΔGv i For individual I i growth rate;

[0128] Npop is the population size;

[0129] Otherwise, the ivy grows upward and climbs, as expressed by the following formula,

[0130]

[0131] Where:

[0132] I Best is the current best individual in the population;

[0133] rand(1,D) is a random real number in the interval [0,1].

[0134] The second aspect of the present invention provides a system for optimizing reactive power compensation of distributed nodes in converter stations using an improved IVY algorithm, and the method for optimizing reactive power compensation of distributed nodes in converter stations using the improved IVY algorithm comprises:

[0135] The power flow calculation module is used to construct the converter model of the converter station and the network equation of the converter station, and solve the power flow calculation model;

[0136] Reactive power compensation module, used to build reactive power compensation configuration model of distributed nodes in DC converter stations;

[0137] Model optimization module, used to optimize the reactive power compensation configuration model and output the optimal reactive power compensation configuration solution under the current power flow;

[0138] The scheme evaluation module is used to evaluate the optimal reactive power compensation configuration scheme, determine whether the maximum number of iterations is met, and detect power flow changes. When the power flow changes, a restart signal is sent to the reactive power compensation module.

[0139] Compared with the prior art, the beneficial effects of the present invention include at least:

[0140] (1) Improving compensation effect: By building an accurate DC converter station distributed node power flow calculation model and solving the model to obtain DC system variables, accurate calculation and adjustment of reactive power compensation can be achieved, thereby improving the compensation effect and meeting the system's demand for reactive power.

[0141] (2) Achieve global optimization: Taking the minimum network loss of the DC converter station as the goal of reactive power compensation configuration, while considering various constraints such as the voltage quality of the distribution nodes, the reactive power output of the converter, the DC converter current, the converter temperature, and the line thermal capacity, a comprehensive objective function is constructed to achieve optimal configuration of the global reactive power distribution.

[0142] (3) Reduce model complexity: Optimize model structure and solution algorithm, reduce computational complexity and time consumption, and improve the real-time and practical performance of the model.

[0143] (4) Comprehensive processing of constraints: Incorporate multiple constraints into the objective function as penalty functions to ensure that the configuration result is optimal while satisfying all constraints.

[0144] (5) Improve algorithm adaptability: Improve the Ivy algorithm so that it can better adapt to DC converter station systems of different scales and operating conditions, and improve the adaptability and robustness of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0145] Figure 1 It is a flow chart of a method for optimizing reactive power compensation of distributed nodes in a DC converter station according to an improved Ivy algorithm provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0146] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0147] When a converter station fails, reactive compensation needs to be implemented to alleviate the compensation effect obtained by the existing compensation method. The present invention first constructs a DC converter station distribution node flow calculation model, analyzes the converter station converter model characteristics, studies the converter station distribution node flow calculation method, and obtains the DC system variables of the DC converter station distribution node, such as voltage, current, power, etc. Then, a reactive compensation configuration model is established, and the minimum network loss of the DC converter station is used as the target of the reactive compensation configuration of the distribution node. The voltage quality of the distribution node, the reactive output of the converter, the DC converter current, the converter temperature, and the line thermal capacity are used as penalty functions to ensure that the reactive compensation configuration scheme can meet the actual needs of the power grid, and the objective function of the reactive compensation configuration model of the distribution node is constructed, and the node power equation constraint is implemented on the objective function. Finally, based on the improved ivy algorithm, the constructed objective function is solved, the optimal reactive compensation configuration scheme is found, and the automatic configuration of reactive compensation of the distribution node is realized. This will help improve the stability and efficiency of the power grid and reduce the operating cost of the power grid.

[0148] like Figure 1 As shown, embodiment 1 of the present invention provides a method for optimizing reactive power compensation of distributed nodes in converter stations using an improved IVY algorithm, comprising the following steps:

[0149] Step 1: construct a power flow calculation model for the distribution nodes of the converter station, including the converter station converter model and the network equation of the converter station.

[0150] In a preferred but non-limiting embodiment of the present invention, step 1 specifically comprises:

[0151] Step 1.1, construct the converter model of the converter station.

[0152] When implementing power flow calculation at the distributed nodes of the converter station, in order to accurately obtain the relevant parameters of the DC system variables (such as DC voltage, DC current, power, etc.), it is first necessary to establish a suitable power flow calculation mathematical model. This model will take into account the characteristics of the converter in the DC converter station, the topological structure of the DC network, and the interface conditions with the AC system.

[0153] The following are some common assumptions about the converters in a DC converter station that help simplify the model and make it easier to solve:

[0154] (1) The inverter is lossless and highly efficient.

[0155] (2) Idealization of the AC side interface: It is assumed that the interface between the converter and the AC system is ideal, that is, lossless and stable in performance.

[0156] (3) Ignore the excitation current.

[0157] (4) Idealization of the AC side interface: It is assumed that the interface between the converter and the AC system (such as transformer and filter) is ideal, that is, lossless and stable. This helps to focus on the power flow calculation on the DC side.

[0158] (5) System stability: Assume that the system is in a stable state, that is, the voltage, current, and power parameters of each node are constant and do not change over time.

[0159] Further optimization is carried out to construct a converter model for the five control modes of the converter station, which is expressed as follows:

[0160]

[0161] Where:

[0162] I b is the converter current of the DC converter station;

[0163] It refers to the distributed node current;

[0164] U b Indicates the converter voltage;

[0165] It refers to the voltage of the distribution node;

[0166] D r It refers to the transformer ratio of the converter;

[0167] Distribution node control angle;

[0168] λ b is the converter control angle;

[0169] is the transformer ratio of the distribution node;

[0170] Refers to the distributed node power.

[0171] Step 1.2, construct the network equation of the converter station, expressed as the following formula:

[0172] B(U w ,A)=0 (2)

[0173] Where:

[0174] B represents the converter station control coefficient;

[0175] U w Represents the bus voltage of the DC converter station;

[0176] A represents the DC system variable.

[0177] Step 1.3, the converter equations and the converter station DC network equations constitute the power flow calculation model, which is expressed as the following formula:

[0178]

[0179] Where:

[0180] R h is the power baseline value for the entire network;

[0181] B(1) is the constant transformation ratio control coefficient of the converter station transformer;

[0182] B(2) is the fixed control angle control coefficient of the converter station;

[0183] B(3) is the constant power control coefficient of the converter station;

[0184] B(4) is the constant current control coefficient of the converter station;

[0185] B(5) is the constant voltage control coefficient of the converter station;

[0186] M a Represents the number of bridges in series;

[0187] L is the repeated part in the commutation;

[0188] f β is the voltage ratio;

[0189] is the inverter angle;

[0190] Y f is the reactance;

[0191] F(U b ,I b ) refers to U b with I b The loop function.

[0192] Step 2: define the DC system variables, solve the power flow calculation model, and obtain the DC system variables of the DC converter station distribution nodes, such as voltage and current.

[0193] Preferably, the power flow calculation model obtained in step 1 is solved to obtain the DC system variables of the distribution nodes, and the distributed node power flow calculation of the converter station is completed. The calculation result is expressed by the following formula:

[0194]

[0195] Step 3: Based on the DC system variables, namely the power flow calculation results of the converter station distribution nodes, a reactive compensation configuration model for the DC converter station distribution nodes is constructed.

[0196] Specifically, the reactive power compensation configuration of distributed nodes is taken as the optimization object, the minimum network loss of the converter station is taken as the goal, and a penalty function is introduced to limit the out-of-bounds of constraint variables. The objective function (fitness function) of the reactive power compensation configuration model of distributed nodes of the DC converter station is constructed, and the node power equation constraint is implemented on the objective function.

[0197] In a preferred but non-limiting embodiment of the present invention, step 3 specifically comprises:

[0198] Step 3.1, set the goal to minimize the network loss of the DC converter station. The voltage quality of the distribution node, the reactive output of the converter, the DC converter current, the converter temperature and the line thermal capacity are taken as constraint variables, and specific constraints are added to the objective function in the form of penalty functions, including the DC current out-of-bounds penalty function, the converter reactive output out-of-bounds penalty function, the distribution node voltage out-of-bounds penalty function, the converter temperature out-of-bounds penalty function and the line thermal capacity out-of-bounds penalty function.

[0199] In a further preferred but non-limiting embodiment of the present invention, step 3.1 specifically comprises:

[0200] Step 3.1.1, construct the corresponding out-of-bounds penalty function of DC current;

[0201]

[0202] Where:

[0203] ε1 refers to the out-of-bounds penalty coefficient corresponding to the DC current;

[0204] H x is the number of DC network nodes;

[0205] I m It refers to the current passing through the DC network node of the converter;

[0206] I max I m The maximum value of

[0207] I min I m The minimum value of .

[0208] Step 3.1.2, constructing the corresponding out-of-bounds penalty function of the reactive output of the converter;

[0209]

[0210] Where:

[0211] ε2 is the corresponding out-of-bounds penalty coefficient of the converter reactive output;

[0212] H y It refers to the total number of inverters;

[0213] E j It refers to the reactive output of the converter;

[0214] E jmax Represents E j Maximum value;

[0215] E jmin Refers to E j Minimum value.

[0216] Step 3.1.3, construct the corresponding out-of-bounds penalty function of the distributed node voltage;

[0217]

[0218] Where:

[0219] ε3 refers to the corresponding out-of-bounds penalty coefficient of the distribution node voltage;

[0220] H j Refers to the total number of distribution nodes;

[0221] U i It refers to the voltage of the distribution node;

[0222] U imax Refers to the maximum value of the voltage at the distribution node;

[0223] U imin Refers to the minimum value of the distribution node voltage.

[0224] Step 3.1.4, construct the inverter temperature out-of-bounds penalty function. The inverter temperature is a key limiting factor. A penalty function can be added to punish excessively high temperatures:

[0225]

[0226] Where:

[0227] ε4 is the corresponding out-of-bounds penalty coefficient of the converter temperature;

[0228] Where T j is the temperature of converter j;

[0229] T max,j is its maximum allowable temperature.

[0230] Step 3.1.5, construct the penalty function for the case of approaching or exceeding the line thermal capacity:

[0231]

[0232] Where:

[0233] ε5 is the corresponding out-of-bounds penalty coefficient of the line thermal capacity;

[0234] I ij is the actual current of line ij;

[0235] I thernal,ij is the thermal capacity current of line ij;

[0236] Number of lines H x is the number of DC network nodes.

[0237] Step 3.1.6, add the active power loss of the DC converter station distribution nodes and the reactive power loss of the DC converter station converter to the objective function. The objective function is expressed as follows:

[0238] minX=f1+f2+f3+f4+f5+Q1+Q2 (10)

[0239] Where:

[0240] minX refers to the minimum network loss of the converter station; Q1 refers to the active power loss of the distributed nodes of the DC converter station; Q2 refers to the reactive power loss of the converter in the DC converter station.

[0241] Step 3.2, impose node power equation constraints on the objective function to ensure that the configuration scheme meets the physical and operational constraints of the power system. The power equation constraints are formulated as follows:

[0242]

[0243] Where:

[0244] P i1 Refers to the corresponding active power output of the converter node;

[0245] P i2 Represents the corresponding active load power of the distribution node;

[0246] U m is the converter voltage;

[0247] U n Refers to the average voltage of the distribution node;

[0248] H j is the total number of distribution nodes;

[0249] Q j1 Refers to the actual reactive power compensation capacity of the distribution node;

[0250] W mn It refers to the conductance phase angle difference between the converter node and the distribution node;

[0251] Q i1 Represents the actual reactive power output of the converter node;

[0252] Represents the voltage phase angle difference between the distribution node and the converter node;

[0253] P i3 Represents the actual active input of the DC network node;

[0254] S mn It refers to the susceptance phase angle difference between the distribution node and the converter node;

[0255] Q i3 Refers to the actual reactive power input of the DC network node;

[0256] m refers to a certain converter;

[0257] Q i2 Indicates the actual reactive load power of the distribution node;

[0258] H y is the total number of converters.

[0259] Step 4, using the improved IVY algorithm to optimize the reactive compensation configuration model of the distributed nodes of the DC converter station constructed in step 3, solving the objective function, and obtaining the optimal reactive compensation configuration scheme under the current power flow. The improved IVY algorithm includes: taking the reactive compensation configuration scheme as an individual of the population, randomly generating the initial position of the population in the search space, and dynamically adjusting the growth rate according to the fitness distribution of the current population and the situation of the search space.

[0260] IVY refers to the ivy plant, and this algorithm mainly simulates the different life stages of ivy, including growth, ascent and spread in the ivy plant community, in order to find the optimal solution.

[0261] In a preferred but non-limiting embodiment of the present invention, the IVY algorithm is improved according to the characteristics of the reactive power compensation optimization problem of the converter station distribution nodes to improve the solution efficiency and accuracy. Step 4 specifically includes:

[0262] Step 4.1, using the reactive power compensation configuration scheme as an individual in the IVY algorithm, multiple individuals form an IVY population, and the upper and lower limits of the constituent elements of each individual form a search space to initialize the IVY population.

[0263] Further preferably, the initialization formula of the original algorithm is improved, and the initial position of the IVY population in the search space is randomly determined using formula (12), so that the population can be more evenly distributed in the search space, thereby increasing the possibility of finding the global optimal solution.

[0264] I i =I min +rand(1,D)⊙(I max -I min ),i=1,…,Npop (12)

[0265] Where:

[0266] rand(1,D) represents a vector of uniformly distributed random numbers in the interval [0,1] with dimension D, where D has the same number of elements as the population I;

[0267] I max and I min are the upper and lower bounds of the search space respectively;

[0268] ⊙ represents the Hadamard product of two vectors, also called element-wise product.

[0269] Among them, the constituent elements of population I are expressed by the following formula:

[0270] I=[V A1 …V AN |Q B1 …Q BN |U C1 …U CN ]

[0271] Where:

[0272] V AN Indicates the corresponding tap position of the Nth transformer;

[0273] Q BNRepresents the actual reactive power compensation amount of the Nth reactive power compensation point;

[0274] U CN represents the corresponding voltage of the Nth generator node.

[0275] It can be understood that at this time D=3N.

[0276]

[0277] Where:

[0278] V ANmax 、V ANmin are the maximum and minimum transformer tap positions.

[0279] Q BNmax , Q BNmin The maximum and minimum reactive power compensation

[0280] U CNmax , U CNmin are the maximum and minimum terminal voltages.

[0281] Decompose the above formula to obtain the control variable represented by the per-unit value, as shown in the following formula:

[0282]

[0283] Where:

[0284] ΔW E Indicates the switching capacity of each group of capacitors; W E Represents the total switching capacity of the capacitor; U P Indicates the actual voltage value at the generator end; T P Represents the actual change of the transformer; ΔT p Indicates the actual change of each gear of the transformer.

[0285] The determined control variables are mapped to each individual in the IVY population. Each individual will represent a specific reactive power compensation configuration scheme, and its basic solution vector is composed of the values ​​of the control variables.

[0286] Step 4.2, construct a coordinated and orderly population growth process in the improved IVY algorithm.

[0287] In the original algorithm, the growth of the population is based on a certain growth rate. The present invention introduces a dynamic growth rate adjustment mechanism, which dynamically adjusts the growth rate according to the fitness distribution of the current population (the distribution of the objective functions of each reactive compensation configuration scheme) and the conditions of the search space (the upper and lower limits of the control variables), so that the population can more effectively explore the potential optimal solution area during the search process.

[0288] Let G vis the growth rate, based on data-intensive experiments and simulations, individual I i The growth rate Gv i (t) is calculated according to the following difference equation,

[0289]

[0290] Where:

[0291] Vector ΔGv i (t) and ΔGvi(t+1) represent the growth rates of the discrete-time system (at time t and time t+1);

[0292] rand is a random real number in the interval [0,1] (i.e. rand∈U[0,1]);

[0293] rand 2 is a random number of a random variable whose probability density function is equal to 1 / (2√x));

[0294] N(1,D) represents a random vector of dimension D, which is a random number in a standard Gaussian (normal) distribution;

[0295] is the mean of the objective function values ​​of all individuals in the population;

[0296] is a distribution vector, each element of which is the mean of all individuals corresponding to the same constituent element in the population;

[0297] For point division, is the element-wise division of vector u by vector v.

[0298] It can be understood that by using the difference equation to dynamically adjust the population growth rate in the IVY algorithm, taking into account the influence of adjacent individuals in the population, and adding the distribution vector as one of the terms in the difference equation, the situation of the solution around each individual in the randomly generated initial population can be deeply explored, which is conducive to finding the global optimal solution more deeply and quickly.

[0299] Step 4.3, construct the leaf formation stage in the improved IVY algorithm, that is, the expansion stage.

[0300] Further preferably, individual I i Using Individual I ii Climb and move logically along the direction of the light source. The expansion phase is expressed by the following formula:

[0301]

[0302] Update current individual I i The growth rate ΔGv i, expressed as the following formula,

[0303]

[0304] Where:

[0305] |N(1,D)| is a vector whose components are the absolute values ​​of the components of vector N(1,D);

[0306] Iter is the number of iterations.

[0307] Step 4.4, construct the young leaf formation stage in the improved IVY algorithm, that is, the climbing stage.

[0308] In individual I i By roaming globally through the search space to the nearest and most important neighbor I ii After the stage, the next stage is entered, individual I i Try to directly follow the current best individual I of the entire population Best , which is equivalent to Best Look for a better optimal solution around. The climbing stage is expressed by the following formula:

[0309]

[0310] Update current individual Growth rate It is expressed as the following formula:

[0311]

[0312] Step 4.5, in order to simulate the alternation of steps 4.3 and 4.4 in the life of the ivy tree, the survivor selection process in the improved IVY algorithm is constructed.

[0313] Specifically, in the improved IVY algorithm, the parameter β is set to (2+rand) / 2, and the objective function values ​​of the current individual and the current optimal individual are compared, and the expansion or climbing decision is made based on the comparison result. i The objective function value f(I i ) is less than f(I best ), the ivy expands the width of branches and leaves according to formula (14); otherwise, the ivy grows upward and climbs according to formula (16).

[0314] It can be understood that compared with other optimization algorithms, the improved IVY algorithm can more accurately determine the reactive power compensation capacity and location, thereby significantly improving the power factor of the power system. This not only helps to reduce line losses, but also improves the operating efficiency and stability of the power grid.

[0315] Through reasonable reactive power compensation, the improved IVY algorithm can significantly reduce the degree of voltage fluctuation and voltage sag, and improve the voltage stability of the power system. This is of great significance for protecting power equipment and ensuring the safe operation of the power grid.

[0316] During the search process, the IVY algorithm can balance the capabilities of local search and global search to avoid falling into the local optimal solution. This enables the algorithm to find the global optimal solution or a solution close to the global optimal solution when solving the reactive power compensation optimization problem.

[0317] Step 5: Evaluate the optimal reactive power compensation configuration solution obtained in step 4 to verify whether it meets all constraints and evaluate its improvement effect on the network loss of the DC converter station.

[0318] Step 6, setting the restart and termination of the improved IVY algorithm, including: judging whether the maximum number of iterations is met, if so, the algorithm is terminated, judging whether the power flow changes, if so, the algorithm is restarted, and returning to step 3.

[0319] Specifically, determine whether the number of iterations reaches the preset maximum number of iterations or whether the calculated fitness value (objective function value) reaches the set lower limit. If both conditions are not met, add one to the number of iterations and re-optimize the reactive compensation configuration; if any one of the conditions or both conditions are met, end the iteration, select the final optimization plan, and output the result.

[0320] When the power flow changes, the reactive compensation also changes accordingly. At this time, it is necessary to restart the IVY algorithm to make the reactive compensation of the distributed nodes of the converter station work again near the optimal network loss, thereby realizing the automatic configuration optimization of the reactive compensation of the distributed nodes of the DC converter station based on the IVY algorithm.

[0321] Embodiment 2 of the present invention provides a system for optimizing reactive power compensation of distributed nodes of converter stations using an improved IVY algorithm, and the method for optimizing reactive power compensation of distributed nodes of converter stations using the improved IVY algorithm described in Embodiment 1 includes:

[0322] The power flow calculation module is used to construct the converter model of the converter station and the network equation of the converter station, and solve the power flow calculation model;

[0323] Reactive power compensation module, used to build reactive power compensation configuration model of distributed nodes in DC converter stations;

[0324] Model optimization module, used to optimize the reactive power compensation configuration model and output the optimal reactive power compensation configuration solution under the current power flow;

[0325] The scheme evaluation module is used to evaluate the optimal reactive power compensation configuration scheme, determine whether the maximum number of iterations is met, and detect power flow changes. When the power flow changes, a restart signal is sent to the reactive power compensation module.

[0326] Compared with the prior art, the beneficial effects of the present invention include at least:

[0327] (1) Improving compensation effect: By building an accurate DC converter station distributed node power flow calculation model and solving the model to obtain DC system variables, accurate calculation and adjustment of reactive power compensation can be achieved, thereby improving the compensation effect and meeting the system's demand for reactive power.

[0328] (2) Achieve global optimization: Taking the minimum network loss of the DC converter station as the goal of reactive power compensation configuration, while considering various constraints such as the voltage quality of the distribution nodes, the reactive power output of the converter, the DC converter current, the converter temperature, and the line thermal capacity, a comprehensive objective function is constructed to achieve optimal configuration of the global reactive power distribution.

[0329] (3) Reduce model complexity: Optimize model structure and solution algorithm, reduce computational complexity and time consumption, and improve the real-time and practical performance of the model.

[0330] (4) Comprehensive processing of constraints: Incorporate multiple constraints into the objective function as penalty functions to ensure that the configuration result is optimal while satisfying all constraints.

[0331] (5) Improve algorithm adaptability: Improve the Ivy algorithm so that it can better adapt to DC converter station systems of different scales and operating conditions, and improve the adaptability and robustness of the algorithm.

[0332] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0333] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing reactive power compensation of distributed nodes in converter stations based on an improved IVY algorithm, characterized in that: The following steps are involved: Step 1, constructing a power flow calculation model of the converter station distribution nodes, including a converter station converter model and a network equation of the converter station; Step 2, solving the power flow calculation model to obtain the DC system variables of the converter station distribution nodes; Step 3: According to the DC system variables, the reactive power compensation configuration of the distributed nodes is optimized, the minimum network loss of the converter station is optimized, and the penalty function is introduced to limit the constraint variable from crossing the boundary, the objective function is constructed, and the node power equation constraint is implemented on the objective function to obtain the reactive power compensation configuration model of the distributed nodes of the converter station; Step 4, using the improved IVY algorithm to optimize the reactive power compensation configuration model of the distributed nodes of the DC converter station, and obtain the optimal reactive power compensation configuration scheme under the current power flow; Step 5: Evaluate the optimal reactive power compensation configuration scheme to verify whether it meets all constraints; Step 6, determine whether the maximum number of iterations is met, if so, end the step; detect the flow change in real time, and return to step 3 when the flow changes.

2. The method for optimizing reactive power compensation of distributed nodes of converter stations using the improved IVY algorithm according to claim 1, characterized in that: In step 1, the converter model of the converter station is expressed by the following formula: Where: I b is the converter current of the DC converter station; It refers to the distributed node current; U b Indicates the converter voltage; It refers to the voltage of the distribution node; D r It refers to the transformer ratio of the converter; Distribution node control angle; λ b is the converter control angle; is the transformer ratio of the distribution node; Refers to the distributed node power; The network equation of the converter station is expressed as follows: B(U w ,A)=0 Where: B represents the converter station control coefficient; U w Represents the bus voltage of the DC converter station; A represents the DC system variable.

3. The method for optimizing reactive power compensation of converter station distribution nodes using the improved IVY algorithm according to claim 1, characterized in that: In step 3, the constraint variables include: distribution node voltage, converter reactive output, DC converter current, converter temperature and line thermal capacity; the penalty functions include: DC current out-of-bounds penalty function, converter reactive output out-of-bounds penalty function, distribution node voltage out-of-bounds penalty function, converter temperature out-of-bounds penalty function and line thermal capacity out-of-bounds penalty function.

4. The method for optimizing reactive power compensation of distributed nodes of converter stations using the improved IVY algorithm according to claim 3, characterized in that: The DC current out-of-bounds penalty function is expressed as follows: Where: ε1 refers to the out-of-bounds penalty coefficient corresponding to the DC current; H x is the number of DC network nodes; I m It refers to the current passing through the DC network node of the converter; I max Indicates I m The maximum value of I min Indicates I m The minimum value of The inverter reactive output out-of-bounds penalty function is expressed as follows: Where: ε2 is the corresponding penalty coefficient for the reactive output of the converter; H y It refers to the total number of inverters; E j It refers to the reactive output of the converter; E jmax Represents E j Maximum value; E jmin Refers to E j Minimum value; The penalty function for voltage out-of-bounds distribution node is expressed as follows: Where: ε3 refers to the corresponding out-of-bounds penalty coefficient of the distribution node voltage; H j Refers to the total number of distribution nodes; U i It refers to the voltage of the distribution node; U imax Refers to the maximum value of the voltage at the distribution node; U imin Refers to the minimum value of the voltage at the distribution node; The inverter temperature out-of-bounds penalty function is expressed as follows: Where: ε4 is the corresponding out-of-bounds penalty coefficient of the converter temperature; T j is the temperature of converter j; T max,j is its maximum allowable temperature; The line thermal capacity out-of-bounds penalty function is expressed as follows: Where: ε5 is the corresponding out-of-bounds penalty coefficient of the line thermal capacity; I ij is the actual current of line ij; I thernal,ij is the thermal capacity current of line ij; Number of lines H x is the number of DC network nodes.

5. The method for optimizing reactive power compensation of converter station distribution nodes using the improved IVY algorithm according to claim 4, characterized in that: In step 3, the objective function is expressed as follows: minX=f1+f2+f3+f4+f5+Q1+Q2 Where: minX refers to the minimum network loss of the converter station; Q1 refers to the active power loss of the distributed nodes of the DC converter station; Q2 refers to the reactive power loss of the DC converter station converter.

6. The method for optimizing reactive power compensation of distributed nodes of converter stations using the improved IVY algorithm according to claim 1, characterized in that: In step 3, the node power equation constraint is expressed as follows: Where: P i1 is the corresponding active power output of the converter node; P i2 is the corresponding active load power of the distribution node; P i3 is the actual active input of the DC network node; U m is the converter voltage; H j is the total number of distribution nodes; U n is the average voltage of the distribution node; W mn It refers to the conductance phase angle difference between the converter node and the distribution node; is the voltage phase angle difference between the distribution node and the converter node; S mn is the susceptance phase angle difference between the distribution node and the converter node; Q i1 is the actual reactive power output of the converter node; Q j1 is the actual reactive power compensation capacity of the distribution node; Q i2 is the actual reactive load power of the distribution node; Q i3 is the actual reactive power input of the DC network node; H y is the total number of converters.

7. The method for optimizing reactive power compensation of distributed nodes of converter stations using the improved IVY algorithm according to claim 1, characterized in that: In step 4, the optimization of the reactive power compensation configuration model of the distributed nodes of the DC converter station by using the improved IVY algorithm also includes: The initial position of the IVY population in the search space is determined by the following formula: I i =I min +rand(1,D)⊙(I max -I min ),i=1,…,Npop Where: rand(1,D) represents a vector of dimension D of uniformly distributed random numbers in the interval [0,1]; I max and I min are the upper and lower bounds of the search space respectively; ⊙ represents the element-wise product; Npop is the population size.

8. The method for optimizing reactive power compensation of converter station distribution nodes using the improved IVY algorithm according to claim 1, characterized in that: In step 4, the optimization of the reactive power compensation configuration model of the distributed nodes of the DC converter station by using the improved IVY algorithm also includes: The growth rate is calculated according to the following difference equation, Where: Vector ΔGv i (t) and ΔGvi(t+1) represent the growth rates of the discrete-time system at time t and time t+1; rand is a random real number in the interval [0,1]; rand 2 is a random number of a random variable whose probability density function is equal to N(1,D) represents a random vector with dimension D, which is a random number in the standard normal distribution; is the mean of the objective function values ​​of all individuals in the population; is a distribution vector, each element of which is the mean of all individuals corresponding to the same constituent element in the population; For point division, is the element-wise division of vector u by vector v.

9. The method for optimizing reactive power compensation of distributed nodes of converter stations using the improved IVY algorithm according to claim 1, characterized in that: In step 4, the optimization of the reactive power compensation configuration model of the distributed nodes of the DC converter station by using the improved IVY algorithm also includes: Perform survivor selection, when individual I i The objective function value is less than the current optimal individual I best When the objective function value is β times, β = (2 + rand) / 2, rand is a random real number in the interval [0,1]. The width of ivy branches and leaves is expressed as follows: Where: Update results for the current individual; I ii For individual I i Neighboring individuals; N(1,D) represents a random vector of dimension D, whose components are random numbers from the standard normal distribution; |N(1,D)| is a vector of dimension D consisting of the absolute values ​​of the components of N(1,D); ΔGv i For individual I i growth rate; Npop is the population size; Otherwise, the ivy grows upward and climbs, as expressed by the following formula, Where: I Best is the current best individual in the population; rand(1,D) is a random real number in the interval [0,1].

10. A system for optimizing reactive power compensation of distributed nodes of converter stations using an improved IVY algorithm, which runs the method for optimizing reactive power compensation of distributed nodes of converter stations using an improved IVY algorithm as claimed in any one of claims 1 to 9, characterized in that: include: The power flow calculation module is used to construct the converter model of the converter station and the network equation of the converter station, and solve the power flow calculation model; Reactive power compensation module, used to build reactive power compensation configuration model of distributed nodes in DC converter stations; Model optimization module, used to optimize the reactive power compensation configuration model and output the optimal reactive power compensation configuration solution under the current power flow; The scheme evaluation module is used to evaluate the optimal reactive power compensation configuration scheme, determine whether the maximum number of iterations is met, and detect power flow changes. When the power flow changes, a restart signal is sent to the reactive power compensation module.

Images