An automatic voltage control method and system based on an improved ant colony algorithm

Abstract

An automatic voltage control method and system based on an improved ant colony algorithm is disclosed. The method includes: collecting distribution network parameters; calculating the sensitivity and voltage deviation of each node and screening nodes to generate a candidate compensation node set; constructing a multi-objective collaborative optimization function with the objectives of minimizing capacitor investment cost, control cost, and total transmission distance, and using the voltage exceedance amplitude and compensation amount deviation of candidate compensation nodes as penalty terms; initializing the parameters of the improved ant colony algorithm; calculating the heuristic pheromone for the compensation capacity level and compensation order at each node, and selecting the compensation capacity level and the next compensation node; locally updating the pheromone concentration after selection, generating a reactive power compensation scheme, selecting the optimal compensation scheme, and updating its global pheromone concentration until iterative convergence, outputting the optimal compensation scheme. This invention enables global optimization configuration of reactive power compensation schemes, thereby effectively reducing operating costs and improving optimization solution efficiency and result reliability.

Description

Technical Field

[0001] This invention belongs to the field of automatic voltage control technology, specifically relating to an automatic voltage control method and system based on an improved ant colony algorithm. Background Technology

[0002] Automatic voltage control is a crucial element in ensuring the safe and stable operation of the power grid. Traditional control methods often rely on linear or nonlinear programming, and their mathematical models are ill-suited to the high precision and real-time requirements of complex power grids. Although improved schemes based on intelligent algorithms (such as traditional ant colony optimization and deep learning) have been introduced, these methods often have a single optimization objective, focusing only on individual indicators such as network losses or costs, failing to achieve synergistic optimization of reactive power compensation, voltage stability, and economic efficiency, resulting in incomplete control performance.

[0003] Currently, existing voltage control methods include: Patent application CN108376986A provides a reactive power voltage control method and device for distribution networks. This method obtains the upper and lower limits of reactive power of each substation in the distribution network and solves the optimal power factor of the substation by obtaining the objective function of network loss. This realizes the power factor control range of each substation to control the switching of reactive power compensation equipment connected to each substation. Patent application CN113541184A discloses a secondary voltage control method for power grids in offshore wind power gathering areas. Based on the existing Coordinated Secondary Voltage Control (CSVC) model, it detects the voltage of the high-voltage side bus of the offshore substation in the offshore wind power gathering area, which can eliminate voltage over-limit at sea. Patent application CN118214003A provides a distributed voltage control method for distribution networks. It uses the network topology information and branch node parameters of the distribution network to construct the optimal power flow model of the distribution network, and then obtains the voltage control value of the corresponding node based on the solution results.

[0004] Existing voltage control methods suffer from significant voltage fluctuations due to changes in local grid load characteristics and high control costs. Even after control, the voltage fluctuation range remains large, making it impossible to strictly stabilize the node voltage within a narrow range, which affects power quality and equipment safety. Summary of the Invention

[0005] To address the technical problems of significant voltage fluctuations and high control costs caused by changes in local power grid load characteristics in existing technologies, this invention provides an automatic voltage control method and system based on an improved ant colony algorithm. The method includes: collecting distribution network parameters, calculating the sensitivity and voltage deviation of each node and screening nodes to generate a candidate compensation node set; constructing a multi-objective collaborative optimization function with the objectives of minimizing capacitor investment costs, control costs, and total transmission distance, and using the voltage exceedance amplitude and compensation amount deviation of candidate compensation nodes as penalty terms; initializing the parameters of the improved ant colony algorithm; calculating the heuristic pheromone for the compensation capacity level and compensation order at each node, and selecting the compensation capacity level and the next compensation node; locally updating the pheromone concentration after selection, generating reactive power compensation schemes, selecting the optimal compensation scheme, and updating its global pheromone concentration until iterative convergence, outputting the optimal compensation scheme. This invention enables global optimization configuration of reactive power compensation schemes, thereby effectively reducing operating costs and improving optimization solution efficiency and result reliability.

[0006] The present invention adopts the following technical solution: A first aspect of the present invention provides an automatic voltage control method based on an improved ant colony algorithm, comprising: S1. Real-time acquisition of voltage, active and reactive power, network topology and line parameters of each node in the distribution network; S2. Based on the collected voltage, active power, and reactive power of each node in the distribution network, calculate the sensitivity and voltage deviation of each node and screen the nodes to generate a set of candidate compensation nodes. Calculate the corresponding preliminary total compensation amount. Based on the network topology and the set of candidate compensation nodes, calculate the total transmission distance. With the goal of minimizing the capacitor investment cost, control cost, and total transmission distance required for reactive power compensation, and with the voltage over-limit amplitude and compensation amount deviation of the candidate compensation nodes as penalty terms, construct a multi-objective collaborative optimization function. S3. Define the compensation capacity level for each candidate compensation node and initialize the parameters of the improved ant colony algorithm. During the iteration process, calculate the heuristic pheromone for the compensation capacity level and compensation order under each candidate compensation node, determine the level selection probability and node selection probability, and select the compensation capacity level and the next compensation node. After selection, locally update the pheromone concentration. After traversing all candidate compensation nodes, generate a reactive power compensation scheme. Select the optimal compensation scheme based on the multi-objective collaborative optimization function value, update its global pheromone concentration, and continue iteratively until convergence, outputting the optimal compensation scheme.

[0007] Preferably, the process of generating the candidate compensation node set in S2 is as follows: For each node in the distribution network, a predefined reactive power disturbance is applied to the corresponding node, and the voltage change of the corresponding node before and after the disturbance is measured. The ratio of the voltage change to the reactive power disturbance is used as the sensitivity of the corresponding node voltage to reactive power injection. The voltage deviation of the corresponding node is obtained by subtracting the voltage of the rated voltage from the voltage of the corresponding node. For each node in the distribution network, if the corresponding sensitivity is not less than a predefined sensitivity threshold and the voltage deviation is not less than a predefined voltage deviation threshold, then the corresponding node is a candidate compensation node; otherwise, it is not a candidate compensation node. Based on all candidate compensation nodes, a set of candidate compensation nodes is generated.

[0008] Preferably, the process of calculating the preliminary total compensation and total transmission distance in S2 is as follows: For each candidate compensation node in the candidate compensation node set, the voltage deviation of the corresponding node is divided by the corresponding sensitivity to obtain the theoretical reactive power compensation amount of the corresponding candidate compensation node; based on the upper and lower limit constraints of the capacity of the reactive power compensation equipment, the theoretical reactive power compensation amount is limited to the upper and lower limit range to generate the preliminary reactive power compensation amount; the preliminary reactive power compensation amounts of all candidate compensation nodes in the candidate compensation node set are summed to obtain the preliminary total compensation amount. Based on the set of candidate compensation nodes, the compensation order of the candidate compensation nodes is set; the electrical distances between the candidate compensation nodes under the compensation order are accumulated to calculate the total transmission distance.

[0009] Preferably, the process of constructing the multi-objective collaborative optimization function in S2 is as follows: The cost of the compensation equipment is obtained by multiplying the number of capacitor banks invested in each candidate compensation node in the candidate compensation node set by the price of the corresponding capacitor and summing them up. The difference between the active power loss of the entire distribution network after reactive power compensation of the candidate compensation node and the active power loss of the entire network before compensation is calculated. The maximum recovery life of the capacitor is multiplied by the difference between the annual maximum load utilization hours, the electricity price and the active power loss in sequence. Based on the predefined upper voltage limit constraint, candidate compensation nodes whose voltage exceeds the predetermined upper and lower voltage limits are selected from the candidate compensation node set to generate a voltage over-limit node set; the voltage over-limit value of each over-limit node in the voltage over-limit node set is calculated as the corresponding voltage over-limit amplitude; the voltage over-limit amplitudes of all over-limit nodes in the voltage over-limit node set are summed and multiplied by a predefined penalty factor as the voltage over-limit penalty; the ratio of the difference between the actual reactive power compensation amount and the initial reactive power compensation amount of each candidate compensation node to the initial total compensation amount is calculated, summed, and multiplied by a predefined compensation offset penalty coefficient as the compensation amount offset penalty term; By adding the capacitor investment cost, control cost, total transmission distance, voltage over-limit penalty, and compensation offset penalty, a multi-objective collaborative optimization function is obtained.

[0010] Preferably, in S3, the pheromone concentration, pheromone importance, heuristic pheromone importance, and adaptive pheromone evaporation coefficient are initialized; the total number of candidate compensation nodes in the candidate compensation node set is taken as the ant colony size, and each candidate compensation node is taken as the starting point of the corresponding ant; based on the candidate compensation node set and the predefined upper and lower limits of the reactive power compensation equipment capacity of each candidate compensation node, the compensation capacity level of each candidate compensation node is defined. After each ant completes the selection of a candidate compensation node and compensation capacity level, the pheromone concentration is locally updated. After each ant visits all candidate compensation nodes, a corresponding reactive power compensation scheme is generated, including the compensation order and the compensation capacity of each node. The multi-objective collaborative optimization function value under each reactive power compensation scheme is calculated, and the minimum multi-objective collaborative optimization function value is taken as the optimal function value. The corresponding scheme is taken as the optimal compensation scheme. The pheromone concentration under each compensation order and compensation capacity in the optimal compensation scheme is globally updated. The optimal compensation scheme is output when the maximum number of iterations is reached or the difference between the optimal function values ​​under adjacent iterations is less than a predetermined difference threshold.

[0011] Preferably, the process of calculating the heuristic pheromone, gear selection probability, and node selection probability in S3 is as follows: For each ant, multiply the difference between the compensation capacity level of the currently selected candidate compensation node and the initial reactive power compensation by a predefined scaling factor, add 1, and take the reciprocal as the heuristic pheromone for the corresponding compensation capacity level. For any compensation capacity level of the currently selected candidate compensation node, the pheromone importance is used as the index of the corresponding pheromone concentration to calculate the pheromone term; the heuristic pheromone importance is used as the index of the corresponding heuristic pheromone to calculate the heuristic pheromone term; the product of the pheromone term of the corresponding compensation capacity level and the heuristic pheromone term is used as the numerator, and the sum of the products of the pheromone terms of all compensation capacity levels and the heuristic pheromone terms is used as the denominator to calculate the selection probability of the current candidate compensation node level. Add a predefined constant coefficient to the electrical distance between the currently selected candidate compensation node and the next candidate compensation node, and take the reciprocal as the heuristic pheromone for the corresponding compensation order; replace the heuristic pheromone of the compensation capacity level with the heuristic pheromone of the compensation order, and calculate the node selection probability of the currently selected candidate compensation node.

[0012] Preferably, the process of selecting the compensation capacity level and the next compensation node in S3 is as follows: Generate a random number. When the random number is not greater than the set random ratio threshold, take the compensation capacity level corresponding to the maximum value of the product of the pheromone item and the heuristic pheromone item as the compensation capacity level of the currently selected candidate compensation node. When the random number is not greater than the set random ratio threshold, randomly select the compensation capacity level of the currently selected candidate compensation node according to the level selection probability. Based on the node selection probability and the random number, use the same method to arbitrarily select a candidate compensation node from the set of unvisited candidate compensation nodes as the next candidate compensation node.

[0013] Preferably, the process of locally updating the pheromone concentration in S3 is as follows: The measurement function is obtained by sequentially adding the voltage over-limit amplitude and the electrical distance between the currently selected candidate compensation node and the next candidate compensation node to the compensation capacity level of the currently selected candidate compensation node. Divide the predefined unit constant by the measurement function to obtain the local pheromone concentration increment; Based on the adaptive pheromone evaporation coefficient, the current pheromone concentration and the local pheromone concentration increment are weighted to obtain the locally updated pheromone concentration.

[0014] Preferably, the process of globally updating the pheromone concentration in S3 is as follows: Divide the predefined unit constant by the optimal function value to obtain the global pheromone concentration increment; update the adaptive pheromone evaporation coefficient based on the current iteration number; and weight the pheromone concentration and global pheromone concentration increment for each compensation order and compensation capacity in the optimal compensation scheme based on the updated adaptive pheromone evaporation coefficient to obtain the pheromone concentration corresponding to the next iteration number.

[0015] A second aspect of the present invention provides an automatic voltage control system based on an improved ant colony algorithm, comprising an automatic voltage control method based on an improved ant colony algorithm, including: The data acquisition module collects the voltage, active and reactive power, network topology and line parameters of each node in the distribution network in real time. The objective function construction module calculates the sensitivity and voltage deviation of each node and filters nodes based on the collected voltage, active power and reactive power of each node in the distribution network, generates a set of candidate compensation nodes, and calculates the corresponding preliminary total compensation amount. Based on the network topology and the set of candidate compensation nodes, the total transmission distance is calculated. With the goal of minimizing the capacitor investment cost, control cost and total transmission distance required for reactive power compensation, and with the voltage over-limit amplitude and compensation amount deviation of the candidate compensation nodes as penalty terms, a multi-objective collaborative optimization function is constructed. The reactive power compensation scheme determination module defines the compensation capacity level for each candidate compensation node and initializes the parameters of the improved ant colony algorithm. During the iteration process, it calculates the heuristic pheromone for the compensation capacity level and compensation order under each candidate compensation node, determines the level selection probability and node selection probability, and selects the compensation capacity level and the next compensation node. After selection, it locally updates the pheromone concentration. After traversing all candidate compensation nodes, it generates a reactive power compensation scheme. Based on the multi-objective collaborative optimization function value, it selects the optimal compensation scheme, updates its global pheromone concentration, and continues iteratively until convergence, outputting the optimal compensation scheme.

[0016] Compared with the prior art, the beneficial effects of the present invention include at least the following: 1. This invention applies reactive power disturbances to each node of the distribution network and calculates voltage sensitivity. Combined with voltage deviation for dual screening, it accurately identifies the critical nodes most sensitive to voltage regulation and requiring compensation, avoiding the coarse control problems caused by relying on experience or globally unified regulation in traditional methods. Simultaneously, a voltage exceedance penalty term is introduced during the optimization process to strongly constrain situations where voltage exceeds upper and lower limits, enabling the compensation scheme to proactively avoid voltage exceedance risks during generation. Through the synergistic effect of these mechanisms, the distribution network can maintain voltage stability within the allowable range under different operating conditions, thereby significantly improving the safety and stability of system operation and enhancing its adaptability to load fluctuations and distributed power source integration.

[0017] 2. This invention constructs a multi-objective collaborative optimization function that includes capacitor investment costs, operation and control costs, and electrical transmission distance. Furthermore, it introduces a compensation offset penalty term to ensure that the actual compensation result closely approximates the theoretical optimal requirement, thereby avoiding resource waste or voltage regulation failure caused by excessive or insufficient reactive power compensation. Simultaneously, by quantifying the change in active power loss across the entire network after compensation and converting it into long-term operating costs, the optimization objective not only focuses on equipment investment but also considers economic benefits within the operating cycle.

[0018] 3. This invention makes multi-dimensional improvements to the traditional ant colony algorithm: On the one hand, by designing a heuristic pheromone that combines the compensation capacity level deviation with the electrical distance between nodes, the initial guidance capability of the search is improved, enabling the algorithm to more quickly approach a reasonable solution space. On the other hand, an adaptive pheromone evaporation coefficient is introduced, allowing the algorithm to dynamically adjust the search intensity at different iteration stages, enhancing exploration capabilities in the early stages and strengthening convergence performance in the later stages, thereby effectively avoiding getting trapped in local optima. Furthermore, through a mechanism combining local and global updates, and by incorporating key factors such as voltage exceedance amplitude and path distance into the pheromone update, the propagation of excellent solutions is strengthened. This not only significantly improves the algorithm's convergence speed but also enhances the global optimality and stability of the solution, making the final reactive power compensation scheme more reliable and practical, suitable for real-time voltage control in complex distribution network environments. Attached Figure Description

[0019] Figure 1 This is a flowchart of an automatic voltage control method based on an improved ant colony algorithm provided by the present invention; Figure 2 This is a graph showing the change of the adaptive pheromone evaporation coefficient with the number of iterations in an embodiment of the present invention; Figure 3 This is a voltage distribution diagram of each node in the power grid after voltage control in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0021] Example 1 Embodiment 1 of the present invention provides an automatic voltage control method based on an improved ant colony algorithm, using a power distribution network system as an application scenario. Figure 1 As shown, the process includes the following steps: S1. Real-time acquisition of voltage, active and reactive power, network topology and line parameters of each node in the distribution network.

[0022] Through sensing devices such as smart meters and phasor measurement units (PMUs), the voltage, active and reactive power of each node in the power grid, as well as the grid topology and branch parameters, are collected in real time. Specifically, the measurement units (PMUs) deployed in each branch collect data in real time on the voltage at both ends of each branch in the distribution network, the current flowing through the branch, and the resistance of the branch.

[0023] S2. Based on the collected voltage, active power, and reactive power of each node in the distribution network, calculate the sensitivity and voltage deviation of each node and screen the nodes to generate a set of candidate compensation nodes. Calculate the corresponding preliminary total compensation amount. Based on the network topology and the set of candidate compensation nodes, calculate the total transmission distance. With the goal of minimizing the capacitor investment cost, control cost, and total transmission distance required for reactive power compensation, and with the voltage over-limit amplitude and compensation amount deviation of the candidate compensation nodes as penalty terms, construct a multi-objective collaborative optimization function.

[0024] For each node in the distribution network, a predefined reactive power disturbance is artificially applied at the corresponding node (by connecting or disconnecting a small set of capacitors / reactors, or changing the setting value of the Static Var Generator (SVG). The voltage change at the corresponding node before and after the disturbance is measured, and the ratio of the voltage change to the reactive power disturbance is used as the sensitivity of the corresponding node's voltage to reactive power injection; specifically expressed as: ; In the formula, The sensitivity of the voltage at node i to reactive power injection at the node is expressed in pu / Mvar. A predefined reactive power disturbance applied at node i; The voltage change at node i before and after the disturbance; The voltage deviation at the corresponding node is calculated by subtracting the voltage at the rated voltage; specifically, it is expressed as follows: ; In the formula, Let be the voltage deviation at node i; if A value >0 indicates that the current voltage is too low and reactive power injection is needed to raise the voltage; if This indicates that the current voltage is too high and reactive power needs to be absorbed to lower the voltage. The rated voltage is 1.0 pu in this embodiment; The voltage at node i is expressed in per-unit (pu). For each node in the distribution network, if the corresponding sensitivity is not less than a predefined sensitivity threshold and the voltage deviation is not less than a predefined voltage deviation threshold, then the corresponding node is a candidate compensation node; otherwise, it is not a candidate compensation node. Based on all candidate compensation nodes, a set of candidate compensation nodes is generated. In this embodiment, the value range of the predefined sensitivity threshold is 0.01~0.05, and the predefined voltage deviation threshold is 0.02pu.

[0025] As a preferred implementation, after determining the set of candidate compensation nodes, it is necessary to calculate the reactive power compensation amount allocated to each node based on the current voltage deviation of each node and the capacity constraint of the reactive power compensation equipment at that node; the specific process is as follows: For each candidate compensation node in the candidate compensation node set, the voltage deviation of the corresponding node is divided by the corresponding sensitivity to obtain the theoretical reactive power compensation amount for the corresponding candidate compensation node; the specific formula is as follows: ; In the formula, The theoretical reactive power compensation amount of candidate compensation node i is the theoretical reactive power compensation amount required to adjust the voltage of candidate compensation node i to the rated voltage. Setting upper and lower limit constraints on the capacity of reactive power compensation equipment (i.e., the reactive power compensation equipment installed at the candidate compensation node has a limited adjustable range), specifically expressed as: ; In the formula, The actual reactive power compensation amount that candidate compensation node i ultimately invests is the actual reactive power compensation amount, in Mvar. A positive value indicates injected reactive power, and a negative value indicates absorbed reactive power. The minimum reactive power output of the equipment at candidate compensation node i; for capacitor banks that can only be switched on and off, typically... (i.e., completely cut off); for Static Var Generators (SVG). It can be a negative value (for example, -1 Mvar means absorbing 1 Mvar of reactive power). The maximum reactive power output of the equipment at candidate compensation node i; for capacitor banks... This is the total capacity when all capacitors are connected; for SVG, Output power to its rated capacity (e.g., +1 Mvar means 1 Mvar of reactive power is input). Based on the upper and lower limits of the reactive power compensation equipment's capacity, the theoretical reactive power compensation amount is limited within the upper and lower limits to generate the initial reactive power compensation amount; specifically expressed as: ; In the formula, The initial reactive power compensation amount for candidate compensation node i; Let be the clipping function, if < ,but Pick ;like > ,but Pick ;otherwise Pick ; The initial reactive power compensation amounts of all candidate compensation nodes in the candidate compensation node set are summed to obtain the initial total compensation amount; specifically expressed as: ; In the formula, This is the preliminary total compensation amount; This is the set of candidate compensation nodes.

[0026] As a preferred implementation, based on the power grid topology, the resistance and reactance of the line between any two connection nodes are obtained, and the corresponding line impedance magnitude is calculated as the electrical distance between the corresponding connection nodes; that is, d ij =|Z ij |= ; where d ij Z represents the electrical distance between node i and node j. ij R is the impedance of the line between node i and node j; ij X represents the resistance (in Ω) of the line between node i and node j. ij The reactance of the line between node i and node j (unit: Ω); Based on the set of candidate compensation nodes, the compensation order of the candidate compensation nodes is set; the electrical distances between the candidate compensation nodes under the compensation order are accumulated to calculate the total transmission distance; specifically expressed as: ; in, This represents the total transmission distance. path This is the set of compensation paths between candidate compensation nodes given their compensation order; for example, suppose the set of candidate compensation nodes is... The compensation order of the candidate compensation nodes is set as follows: Then the set of compensation paths is .

[0027] As a preferred implementation method, the process of constructing a multi-objective collaborative optimization function is as follows: The cost of the compensation equipment is obtained by multiplying the number of capacitor banks invested in each candidate compensation node in the candidate compensation node set by the price of the corresponding capacitor and then summing them up. Calculate the difference between the total active power loss of the distribution network after reactive power compensation for the candidate compensation nodes and the total active power loss before compensation. Multiply the maximum recovery life of the capacitor by the annual maximum load utilization hours, the electricity price, and the difference in active power loss in sequence to obtain the control cost. Based on the predefined upper limit voltage constraint, candidate compensation nodes whose voltage exceeds the predetermined upper and lower voltage limits are selected from the candidate compensation node set to generate a voltage over-limit node set; the voltage over-limit value of each over-limit node in the voltage over-limit node set is calculated as the corresponding voltage over-limit amplitude; the voltage over-limit amplitudes of all over-limit nodes in the voltage over-limit node set are accumulated and multiplied by a predefined penalty factor to obtain the voltage over-limit penalty. The ratio of the difference between the actual reactive power compensation amount and the initial reactive power compensation amount for each candidate compensation node to the initial total compensation amount is calculated, accumulated, and multiplied by a predefined compensation offset penalty coefficient as the compensation amount offset penalty term. By adding the capacitor investment cost, control cost, total transmission distance, voltage over-limit penalty, and compensation offset penalty, a multi-objective collaborative optimization function is obtained; specifically, it is expressed as: ; in, H For multi-objective collaborative optimization functions; The number of capacitor banks deployed for candidate compensation node i; Let i be the price of each group of capacitors at candidate compensation node i; This refers to the maximum recycling life of the capacitor; This refers to the maximum annual load utilization hours. For electricity price; The active power loss reduced after implementing control is the difference between the active power loss of the entire distribution network after reactive power compensation of the candidate compensation nodes and the active power loss of the entire network before compensation. A predefined penalty factor; The actual reactive power compensation amount for candidate compensation node i; To compensate for the offset penalty coefficient; This is the set of nodes that exceed voltage limits. The voltage over-limit amplitude of node i; The voltage at node i that exceeds the limit; This represents the total transmission distance. This is an indicator function; it returns 1 if the condition within the parentheses is met, and 0 otherwise. Indicates nodes that exceed the limit i The highest voltage; Indicates nodes that exceed the limit i The lowest voltage.

[0028] Optionally, if the candidate compensation node uses a static var generator (SVG), the cost of the compensation equipment is the number of SVGs in all candidate compensation nodes multiplied by the corresponding price. If the candidate compensation node uses capacitor banks and SVG (Static Var Generator), the cost of the compensation equipment is the product of the number of SVGs in the candidate compensation node multiplied by the corresponding price, plus the product of the number of capacitor banks invested and their corresponding prices.

[0029] Since the loss is usually reduced after compensation, the difference is negative; for ease of optimization, its absolute value is taken or the difference is used directly. This embodiment adopts the following form: . This refers to the total active power loss of the distribution network without reactive power compensation. This refers to the total active power loss of the distribution network after applying the current reactive power compensation scheme. When the value is zero, it indicates that the compensation scheme has effectively reduced active power loss. The total active power loss of the distribution network is equal to the sum of the active power losses of all branches. For any branch connecting node i and node j, its active power loss is calculated using the following formula: . The effective value of the current flowing through branch ij The resistance of branch ij.

[0030] S3. Define the compensation capacity level for each candidate compensation node, initialize the pheromone concentration, and take each node in the candidate compensation node set as the starting compensation node. During the iteration process, calculate the heuristic pheromone for the compensation capacity level and compensation order under each candidate compensation node, determine the level selection probability and node selection probability, and select the compensation capacity level and the next compensation node. After selection, locally update the pheromone concentration. After traversing all candidate compensation nodes, generate a reactive power compensation scheme. Select the optimal compensation scheme according to the multi-objective collaborative optimization function value, update its global pheromone concentration, and continue iterating until the iteration converges, outputting the optimal compensation scheme.

[0031] As a preferred implementation method, the multi-objective collaborative optimization function is solved by improving the ant colony algorithm; the specific process is as follows: The total number of candidate compensation nodes in the candidate compensation node set is taken as the ant colony size. Pheromones concentration, pheromone importance, heuristic pheromone importance, and adaptive pheromone evaporation coefficient are initialized. and maximum number of iterations See Figure 2 In this embodiment, the ant colony size is set to 32, the pheromone importance is set to α=1, the heuristic pheromone importance is set to β=5, and the adaptive pheromone volatility coefficient is set to... x corresponds to the iteration number t in this embodiment; each candidate compensation node is taken as the starting point of the corresponding ant; based on the set of candidate compensation nodes and the predefined upper and lower limits of the capacity of the reactive power compensation equipment of each candidate compensation node, the compensation capacity level of each candidate compensation node is defined; For capacitor banks that can only be switched on and off, assume that candidate compensation node i is equipped with Group of capacitors, capacity of each group Mvar, then the compensation capacity level is For SVG, the capacity range of the reactive power compensation equipment of the candidate compensation node is discretized into different capacity levels according to the predefined capacity interval; if the candidate compensation node has both capacitor bank and SVG, they can be combined as an equivalent continuous adjustment range and then discretized.

[0032] During the iteration process, for each ant, the difference between the compensation capacity level of the currently selected candidate compensation node and the initial reactive power compensation amount is multiplied by a predefined scaling factor, then 1 is added, and the reciprocal is taken as the heuristic pheromone for the corresponding compensation capacity level; the specific formula is: ; In the formula, The compensation capacity level for the currently selected candidate compensation node i Heuristic pheromones under the influence of the sun The value is set to the maximum value of 1, and the value decreases as the distance increases, guiding the ants to choose a level closer to the theoretical value. The compensation capacity level for the currently selected candidate compensation node i; This is a predefined scaling factor used to control the sensitivity to compensation bias. For any compensation capacity level of the currently selected candidate compensation node, the pheromone importance is used as the index of the corresponding pheromone concentration to calculate the pheromone term; the heuristic pheromone importance is used as the index of the corresponding heuristic pheromone to calculate the heuristic pheromone term; the product of the pheromone term for the corresponding compensation capacity level and the heuristic pheromone term is used as the numerator, and the sum of the products of the pheromone terms for all compensation capacity levels and the heuristic pheromone terms is used as the denominator to calculate the selection probability of the currently selected candidate compensation node level; the specific formula is: ; In the formula, Let t be the probability of the k-th ant choosing a compensation capacity level at candidate compensation node i under iteration number t; Let pheromone concentration be the pheromone concentration at candidate compensation node i under iteration number t; For pheromone importance; For heuristic pheromone importance; Let i be the set of compensation capacity levels for candidate compensation node i; A random number is generated in the range [0,1]. When the random number is not greater than the set random ratio threshold, the compensation capacity level corresponding to the maximum value of the product of the pheromone item and the heuristic pheromone item is taken as the compensation capacity level of the currently selected candidate compensation node. When the random number is not greater than the set random ratio threshold, the compensation capacity level of the currently selected candidate compensation node is randomly selected according to the level selection probability. Add a predefined constant coefficient to the electrical distance between the currently selected candidate compensation node and the next candidate compensation node, and take the reciprocal as the heuristic pheromone for the corresponding compensation order; replace the heuristic pheromone for the compensation capacity level with the heuristic pheromone for the compensation order, and calculate the node selection probability of the currently selected candidate compensation node; the specific formula is as follows: ; ; In the formula, Select the heuristic pheromone for the next candidate compensation node j under the currently selected candidate compensation node i; This represents a predefined constant coefficient, which is a very small integer (e.g., 10). -6 ), to avoid the denominator being 0; Let be the probability of the k-th ant choosing the next candidate compensation node j at candidate compensation node i under iteration number t; This is the set of unvisited candidate compensation nodes; Based on the node selection probability and random number, the same method is used to randomly select a candidate compensation node from the set of unvisited candidate compensation nodes as the next candidate compensation node. After each ant completes the selection of a candidate compensation node and compensation capacity level, the pheromone concentration is locally updated. After each ant visits all candidate compensation nodes, a corresponding reactive power compensation scheme is generated (including the compensation order and the compensation capacity of each node). The multi-objective collaborative optimization function value under each reactive power compensation scheme is calculated, and the minimum multi-objective collaborative optimization function value is taken as the optimal function value. The corresponding scheme is taken as the optimal compensation scheme. The pheromone concentration under each compensation order and compensation capacity in the optimal compensation scheme is globally updated. The optimal compensation scheme is output when the maximum number of iterations is reached or the difference between the optimal function values ​​under adjacent iterations is less than a predetermined difference threshold.

[0033] As a preferred implementation method, the process of locally updating the pheromone concentration is as follows: Local updates are performed after each ant completes a movement, and the pheromone increment is calculated based on a measurement function M(t). The measurement function is obtained by sequentially adding the compensation capacity level of the currently selected candidate compensation node to the voltage over-limit amplitude and the electrical distance between the currently selected candidate compensation node and the next candidate compensation node; it is defined as: ; in, The compensation capacity level of the candidate compensation node i is injected at the current iteration number t, which is the actual reactive power compensation amount. For voltage exceeding the limit; d ij The electrical distance between the current candidate compensation node i and the next candidate compensation node j; Divide the predefined unit constant by the measurement function to obtain the local pheromone concentration increment; the specific formula is as follows: ; In the formula, This represents the local pheromone concentration increment. For nodes i , For nodes j Q is a unit constant of 1, meaning that the amount of chemical pheromones secreted by each ant along the way is the same. The calculation of pheromone increment is based on a measurement function; Based on the adaptive pheromone evaporation coefficient, the current pheromone concentration and the local pheromone concentration increment are weighted to obtain the locally updated pheromone concentration; specifically expressed as: ; In the formula, This represents the pheromone concentration after a local update. This is the adaptive pheromone evaporation coefficient.

[0034] As a preferred implementation method, the process of globally updating the pheromone concentration is as follows: After the ant colony completes its traversal, it finds the globally optimal path and adjusts the pheromone concentration accordingly. A predefined unit constant is divided by the optimal function value to obtain the global pheromone concentration increment. Based on the current iteration number, the adaptive pheromone volatility coefficient is updated. Based on the updated adaptive pheromone volatility coefficient, the pheromone concentration and the global pheromone concentration increment under each compensation order and compensation capacity in the optimal compensation scheme are weighted to obtain the pheromone concentration corresponding to the next iteration number. Specifically, this is expressed as follows: ; ; In the formula, This represents the global pheromone concentration increment. This represents the optimal function value for the current iteration number. This is the adaptive pheromone evaporation coefficient at the current iteration number t; it is adaptively adjusted according to the iteration number, and its variation pattern is as follows: Figure 2 The negative exponential curve shown demonstrates extensive exploration in the early stages and rapid convergence in the later stages.

[0035] As an optional implementation method, the optimal compensation scheme is evaluated from multiple dimensions: Based on the optimal compensation scheme, the predicted voltage qualification rate of all nodes i in the distribution network is calculated; specifically expressed as: ; In the formula, To predict the voltage qualification rate; Calculate the degree of reduction in distribution network losses after implementing the optimal compensation scheme; specifically: ; In the formula, To reduce network loss; Calculate the cost-benefit ratio of control after implementing the final compensation plan; specifically:

[0036] Among them, C in This represents the total investment cost.

[0037] The optimal compensation scheme is compared with the historical optimal compensation scheme in terms of predicted voltage qualification rate, network loss reduction, and cost-effectiveness, to ensure the reliability and economy of the compensation scheme.

[0038] The optimal compensation scheme is converted into control commands to drive the reactive power compensation equipment; for example... Figure 3 As shown, after control, all node voltages were stabilized within the ideal range of 0.98-1.02 pu, verifying the control effect of the method of the present invention.

[0039] Example 2 Embodiment 2 of the present invention provides an automatic voltage control system based on an improved ant colony algorithm, which executes the automatic voltage control method based on an improved ant colony algorithm described in Embodiment 1, including: The data acquisition module collects the voltage, active and reactive power, network topology and line parameters of each node in the distribution network in real time. The objective function construction module calculates the sensitivity and voltage deviation of each node and filters nodes based on the collected voltage, active power and reactive power of each node in the distribution network, generates a set of candidate compensation nodes, and calculates the corresponding preliminary total compensation amount. Based on the network topology and the set of candidate compensation nodes, the total transmission distance is calculated. With the goal of minimizing the capacitor investment cost, control cost and total transmission distance required for reactive power compensation, and with the voltage over-limit amplitude and compensation amount deviation of the candidate compensation nodes as penalty terms, a multi-objective collaborative optimization function is constructed. The reactive power compensation scheme determination module defines the compensation capacity level for each candidate compensation node and initializes the parameters of the improved ant colony algorithm. During the iteration process, it calculates the heuristic pheromone for the compensation capacity level and compensation order under each candidate compensation node, determines the level selection probability and node selection probability, and selects the compensation capacity level and the next compensation node. After selection, it locally updates the pheromone concentration. After traversing all candidate compensation nodes, it generates a reactive power compensation scheme. Based on the multi-objective collaborative optimization function value, it selects the optimal compensation scheme, updates its global pheromone concentration, and continues iteratively until convergence, outputting the optimal compensation scheme.

[0040] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0041] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0042] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0043] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. An automatic voltage control method based on an improved ant colony algorithm, characterized in that, include: S1. Real-time acquisition of voltage, active and reactive power, network topology and line parameters of each node in the distribution network; S2. Based on the collected voltage, active power and reactive power of each node in the distribution network, calculate the sensitivity and voltage deviation of each node and screen the nodes to generate a set of candidate compensation nodes and calculate the corresponding preliminary total compensation amount. Based on the network topology and combined with the set of candidate compensation nodes, the total transmission distance is calculated. With the goal of minimizing the capacitor investment cost, control cost and total transmission distance required for reactive power compensation, and with the voltage over-limit amplitude and compensation amount deviation of the candidate compensation nodes as penalty terms, a multi-objective collaborative optimization function is constructed. S3. Define the compensation capacity level for each candidate compensation node and initialize the parameters of the improved ant colony algorithm. During the iteration process, the heuristic pheromone for the compensation capacity level and compensation order under each candidate compensation node is calculated to determine the level selection probability and node selection probability, and the compensation capacity level and the next compensation node are selected. After selection, the pheromone concentration is locally updated. After all candidate compensation nodes are traversed, a reactive power compensation scheme is generated. The optimal compensation scheme is selected according to the multi-objective collaborative optimization function value, and its global pheromone concentration is updated until the iteration converges, and the optimal compensation scheme is output.

2. The automatic voltage control method based on the improved ant colony algorithm according to claim 1, characterized in that: The process of generating the candidate compensation node set in S2 is as follows: For each node in the distribution network, a predefined reactive power disturbance is applied to the corresponding node, and the voltage change of the corresponding node before and after the disturbance is measured. The ratio of the voltage change to the reactive power disturbance is used as the sensitivity of the corresponding node voltage to reactive power injection. The voltage deviation of the corresponding node is obtained by subtracting the voltage of the rated voltage from the voltage of the corresponding node. For each node in the distribution network, if the corresponding sensitivity is not less than the predefined sensitivity threshold and the voltage deviation is not less than the predefined voltage deviation threshold, then the corresponding node is a candidate compensation node. Conversely, if not, it is not a candidate compensation node; based on all candidate compensation nodes, a set of candidate compensation nodes is generated.

3. The automatic voltage control method based on the improved ant colony algorithm according to claim 1, characterized in that: The process of calculating the initial total compensation and total transmission distance in S2 is as follows: For each candidate compensation node in the candidate compensation node set, the voltage deviation of the corresponding node is divided by the corresponding sensitivity to obtain the theoretical reactive power compensation amount of the corresponding candidate compensation node. Based on the upper and lower limits of the capacity of the reactive power compensation equipment, the theoretical reactive power compensation amount is limited to the upper and lower limits to generate the preliminary reactive power compensation amount; the preliminary reactive power compensation amounts of all candidate compensation nodes in the candidate compensation node set are summed to obtain the preliminary total compensation amount. Based on the set of candidate compensation nodes, the compensation order of the candidate compensation nodes is set; the electrical distances between the candidate compensation nodes under the compensation order are accumulated to calculate the total transmission distance.

4. The automatic voltage control method based on the improved ant colony algorithm according to claim 1, characterized in that: The process of constructing the multi-objective joint optimization function in S2 is as follows: The cost of the compensation equipment is obtained by multiplying the number of capacitor banks invested in each candidate compensation node in the candidate compensation node set by the price of the corresponding capacitor and summing them up. The difference between the active power loss of the entire distribution network after reactive power compensation of the candidate compensation node and the active power loss of the entire network before compensation is calculated. The maximum recovery life of the capacitor is multiplied by the difference between the annual maximum load utilization hours, the electricity price and the active power loss in sequence. Based on the predefined upper voltage limit constraint, candidate compensation nodes whose voltage exceeds the predetermined upper and lower voltage limits are selected from the candidate compensation node set to generate a voltage over-limit node set; the voltage over-limit value of each over-limit node in the voltage over-limit node set is calculated as the corresponding voltage over-limit amplitude; the voltage over-limit amplitudes of all over-limit nodes in the voltage over-limit node set are summed and multiplied by a predefined penalty factor as the voltage over-limit penalty; the ratio of the difference between the actual reactive power compensation amount and the initial reactive power compensation amount of each candidate compensation node to the initial total compensation amount is calculated, summed, and multiplied by a predefined compensation offset penalty coefficient as the compensation amount offset penalty term; By adding the capacitor investment cost, control cost, total transmission distance, voltage over-limit penalty, and compensation offset penalty, a multi-objective collaborative optimization function is obtained.

5. The automatic voltage control method based on the improved ant colony algorithm according to claim 1, characterized in that: In S3, the pheromone concentration, pheromone importance, heuristic pheromone importance, and adaptive pheromone volatility coefficient are initialized; the total number of candidate compensation nodes in the candidate compensation node set is taken as the ant colony size, and each candidate compensation node is taken as the starting point of the corresponding ant. Based on the set of candidate compensation nodes and the predefined upper and lower limits of the capacity of the reactive power compensation equipment for each candidate compensation node, the compensation capacity level of each candidate compensation node is defined. After each ant completes the selection of a candidate compensation node and compensation capacity level, the pheromone concentration is locally updated. After each ant visits all candidate compensation nodes, a corresponding reactive power compensation scheme is generated, including the compensation order and the compensation capacity of each node. The multi-objective collaborative optimization function value under each reactive power compensation scheme is calculated, and the minimum multi-objective collaborative optimization function value is taken as the optimal function value. The corresponding scheme is taken as the optimal compensation scheme. The pheromone concentration under each compensation order and compensation capacity in the optimal compensation scheme is globally updated. The optimal compensation scheme is output when the maximum number of iterations is reached or the difference between the optimal function values ​​under adjacent iterations is less than a predetermined difference threshold.

6. The automatic voltage control method based on the improved ant colony algorithm according to claim 1, characterized in that: The process of calculating the heuristic pheromone, gear selection probability, and node selection probability in S3 is as follows: For each ant, multiply the difference between the compensation capacity level of the currently selected candidate compensation node and the initial reactive power compensation by a predefined scaling factor, add 1, and take the reciprocal as the heuristic pheromone for the corresponding compensation capacity level. For any compensation capacity level of the currently selected candidate compensation node, the pheromone importance is used as the index of the corresponding pheromone concentration to calculate the pheromone term; The importance of the heuristic pheromone is used as the index of the corresponding heuristic pheromone to calculate the heuristic pheromone term; the product of the pheromone term corresponding to the compensation capacity level and the heuristic pheromone term is used as the numerator, and the sum of the products of the pheromone terms of all compensation capacity levels and the heuristic pheromone terms is used as the denominator to calculate the level selection probability of the currently selected candidate compensation node. Add a predefined constant coefficient to the electrical distance between the currently selected candidate compensation node and the next candidate compensation node, and take the reciprocal as the heuristic pheromone for the corresponding compensation order; replace the heuristic pheromone of the compensation capacity level with the heuristic pheromone of the compensation order, and calculate the node selection probability of the currently selected candidate compensation node.

7. An automatic voltage control method based on an improved ant colony algorithm according to claim 1 or 6, characterized in that: The process of selecting the compensation capacity level and the next compensation node in S3 is as follows: Generate random numbers. When the random numbers are not greater than the set random ratio threshold, take the compensation capacity level corresponding to the maximum value of the product of the pheromone item and the heuristic pheromone item as the compensation capacity level of the currently selected candidate compensation node. When the random number is not greater than the set random ratio threshold, the compensation capacity level of the currently selected candidate compensation node is randomly selected according to the level selection probability. Based on the node selection probability and random number, the same method is used to arbitrarily select a candidate compensation node from the set of unvisited candidate compensation nodes as the next candidate compensation node.

8. An automatic voltage control method based on an improved ant colony algorithm according to claim 1 or 5, characterized in that: The process of locally updating pheromone concentration in S3 is as follows: The measurement function is obtained by sequentially adding the voltage over-limit amplitude and the electrical distance between the currently selected candidate compensation node and the next candidate compensation node to the compensation capacity level of the currently selected candidate compensation node. Divide the predefined unit constant by the measurement function to obtain the local pheromone concentration increment; Based on the adaptive pheromone evaporation coefficient, the current pheromone concentration and the local pheromone concentration increment are weighted to obtain the locally updated pheromone concentration.

9. An automatic voltage control method based on an improved ant colony algorithm according to claim 1 or 5, characterized in that: The process of globally updating pheromone concentration in S3 is as follows: Divide the predefined unit constant by the optimal function value to obtain the global pheromone concentration increment; update the adaptive pheromone evaporation coefficient based on the current iteration number; and weight the pheromone concentration and global pheromone concentration increment for each compensation order and compensation capacity in the optimal compensation scheme based on the updated adaptive pheromone evaporation coefficient to obtain the pheromone concentration corresponding to the next iteration number.

10. An automatic voltage control system based on improved ant colony algorithm, using the method of any one of claims 1-9, characterized in that, include: The data acquisition module collects the voltage, active and reactive power, network topology and line parameters of each node in the distribution network in real time. The objective function construction module calculates the sensitivity and voltage deviation of each node and filters nodes based on the collected voltage, active power and reactive power of each node in the distribution network, generates a set of candidate compensation nodes, and calculates the corresponding preliminary total compensation amount. Based on the network topology and combined with the set of candidate compensation nodes, the total transmission distance is calculated. With the goal of minimizing the capacitor investment cost, control cost and total transmission distance required for reactive power compensation, and with the voltage over-limit amplitude and compensation amount deviation of the candidate compensation nodes as penalty terms, a multi-objective collaborative optimization function is constructed. The reactive power compensation scheme determination module defines the compensation capacity level for each candidate compensation node and initializes the parameters of the improved ant colony algorithm. During the iteration process, the heuristic pheromone for the compensation capacity level and compensation order under each candidate compensation node is calculated to determine the level selection probability and node selection probability, and the compensation capacity level and the next compensation node are selected. After selection, the pheromone concentration is locally updated. After all candidate compensation nodes are traversed, a reactive power compensation scheme is generated. The optimal compensation scheme is selected according to the multi-objective collaborative optimization function value, and its global pheromone concentration is updated until the iteration converges, and the optimal compensation scheme is output.

Images