A method for minimizing power loss in a power distribution system based on quantum genetic algorithm to optimize the location and capacity of distributed power sources.

Abstract

This invention provides a distributed generation (DG) optimization configuration method based on a quantum genetic algorithm, aiming to minimize power loss and improve voltage stability in power distribution systems. The method first collects basic data of the power distribution network, including bus numbers, voltage levels, line impedances, bus load demands, and capacity limitations of distributed generation sources. Next, a population containing different DG configuration schemes is initialized using a random number generation algorithm, and the population is iteratively optimized using a quantum genetic algorithm. During the optimization process, quantum selection, crossover, and mutation operations are used to generate a new generation of configuration schemes, and the power loss and voltage level of each scheme are evaluated using a fitness function to gradually approach the optimal solution. After obtaining the optimal configuration scheme, different load conditions (such as load increases or decreases) are simulated to verify and optimize the adaptability of the configuration, ensuring that the system can maintain low power loss and voltage stability under load fluctuations. The final output optimal DG configuration scheme includes bus location, capacity, and optimized power loss and voltage curves, capable of adapting to dynamic load changes. This method can effectively reduce power loss and improve voltage stability in power distribution systems, providing technical support for building efficient and reliable power distribution systems.

Description

Technical Field

[0001] This invention relates to the field of power distribution system optimization, and in particular to a method for optimizing the location and capacity of distributed power sources based on quantum genetic algorithms to minimize power loss in power distribution systems. Background Technology

[0002] With the growth of global energy demand and the increasing awareness of environmental protection, traditional centralized power generation is gradually being supplemented by distributed generation (DG) systems. DG systems, by directly connecting small-scale power generation units (such as solar, battery storage, and wind power) to the distribution network, can reduce transmission losses, improve energy utilization efficiency, and enhance the power supply reliability of the distribution system. However, the configuration of DG and its application in the distribution network still have the following challenges: 1) Improper DG location may lead to an increase in system power loss instead of a decrease; 2) Inappropriate DG location and capacity configuration may cause bus voltage to exceed limits, affecting voltage stability; 3) The load of the distribution system has significant time-varying characteristics, and the DG configuration must be able to adapt to load fluctuations to maintain stable system operation; 4) Optimizing the configuration of distributed power sources involves complex power flow calculations, especially with a huge computational load under multi-load conditions. To address the aforementioned issues, this invention discloses a method for minimizing power loss in a power distribution system by optimizing the location and capacity of distributed generation sources based on a quantum genetic algorithm. This method significantly improves search efficiency and global optimization capabilities through the selection, crossover, and mutation operations of qubits during the optimization process, finding the optimal or near-optimal DG configuration scheme in a shorter time. This effectively reduces power loss in the power distribution system, enhances voltage stability, and adapts to multiple load conditions, thus providing a highly adaptable and efficient distributed generation optimization configuration scheme. Summary of the Invention

[0003] The purpose of this invention is to provide a method for minimizing power loss in a power distribution system based on optimizing the location and capacity of distributed power sources using a quantum genetic algorithm. This method effectively reduces power loss in the power distribution system, stabilizes the voltage profile, and adapts to system changes under different load conditions by optimizing the location and capacity of distributed power sources.

[0004] The technical solution adopted in this invention is a method for minimizing power loss in a power distribution system based on quantum genetic algorithm to optimize the location and capacity of distributed power sources, specifically implemented according to the following steps:

[0005] Step 1: First, collect basic data on the power distribution network, including specific information on buses, lines, and distributed generation equipment (e.g., capacity and location), and initialize the voltage level and power flow of the power system. The purpose of this step is to obtain the power loss and voltage curves of the system in its initial state, which will serve as a reference baseline for optimization.

[0006] Step 2: After obtaining the basic data, a random number generation algorithm is used to generate an initial population containing several different DG (Distributed Generation) configuration schemes. Each configuration scheme is called an "individual," and its genes indicate whether a DG is deployed on each bus. These individuals are substituted into the initial power system model to calculate their corresponding power losses and voltage curves, providing a preliminary assessment for subsequent optimization.

[0007] Step 3: After obtaining the initial population, a quantum genetic algorithm is used to iteratively optimize the population, aiming to minimize power loss and keep the voltage curve within an acceptable range. The optimization process of the quantum genetic algorithm includes quantum selection, quantum crossover, and quantum mutation. First, quantum selection, based on the fitness function, prioritizes individuals with low power loss and stable voltage to enter the next generation. Next, quantum crossover uses quantum superposition and cloning techniques to combine the characteristics of high-fitness individuals to generate new individuals, thereby gradually optimizing the DG configuration. Finally, quantum mutation introduces perturbation through random Pauli gates to increase population diversity and avoid local optima. In each generation, a new generation of configuration is generated through quantum selection, crossover, and mutation, gradually obtaining a better DG location and capacity scheme.

[0008] Step 4: After obtaining the initially optimized configuration, simulate different load conditions of the power distribution system to verify the adaptability of the DG configuration. Specifically, set the load conditions to increase or decrease by a certain percentage (e.g., 25%, 50%), and repeat the above quantum genetic algorithm optimization steps to ensure that the configuration scheme can still achieve minimum power loss and voltage stability under different load conditions. Record the optimal configuration under each load condition and evaluate the changes in the system's power loss and voltage curves under each load condition.

[0009] Step 5: Finally, output the optimal DG location and capacity configuration scheme, including the DG location and capacity of each bus, and the optimized system power loss and voltage curve data. Simultaneously, provide corresponding optimized configurations for different load conditions to ensure that the DG configuration can adapt to dynamic load changes, thereby achieving stable and efficient power supply capabilities. This method implements a dynamic optimization process based on a quantum genetic algorithm, enabling the power distribution system to maintain low power loss and good voltage stability under different load conditions.

[0010] The invention is further characterized by:

[0011] Step 1 is implemented as follows:

[0012] First, data reading and system initialization are performed. By reading the core data of the power distribution network, the system can accurately represent the network structure. Specifically, the data includes: busbar numbers and voltage levels, impedance values ​​(resistance R) of each line. iand reactance X i Bus load (active power P) i and no Q i This includes the rated power and capacity limitations of distributed generation (DG) sources at their installation locations. These data collectively construct the distribution network topology, ensuring accurate bus-to-line connections and providing a foundation for subsequent power flow calculations and DG optimization.

[0013] After completing data reading, the system begins initial power flow calculation, employing a forward-backward sweep method for analysis. This method consists of two phases:

[0014] (1) Backward Sweep: Starting from the farthest bus in the system, the current on each line is calculated progressively towards the power source. Initially, it is assumed that the voltage of each bus is equal to the source voltage. For the load current of each bus, the following formula can be used:

[0015]

[0016] Among them, I ldi P represents the load current of the i-th bus. i and Q i These represent the active and reactive power demands on the bus, V. i This refers to the voltage of the busbars. During the backward scan, the current values ​​in the lines are continuously updated based on the voltage values ​​from the previous iteration until the currents of all buses in the system are initially distributed.

[0017] (2) Forward Sweep: Starting from the power source point and moving outwards, the voltage of each bus is calculated and updated progressively. Given the line impedance Z... i and current I Li Under the given conditions, the voltage of each busbar is calculated using the following formula:

[0018] V i =V u -Z i I Li (2)

[0019] Among them, V i V is the voltage of the i-th bus. u Z is the voltage of the upstream bus. i For line impedance,

[0020] I Li This represents the current in the line. By continuously advancing towards the far-end busbar of the system, the voltage distribution of each busbar can be obtained, completing a preliminary analysis of the voltage profile.

[0021] After completing the initial power flow calculation, the system also needs to verify the initial voltage and power distribution. During this process, it is checked whether the voltage of each bus is within the specified allowable range (e.g., 0.95 to 1.05 pu). The formulas for calculating the upper and lower voltage limits are as follows:

[0022]

[0023] Among them, V base It is the system's reference voltage.

[0024] Simultaneously calculate the total power loss of the system. The formula for calculating the total power loss of the system is as follows:

[0025]

[0026] Among them, P loss I represents the total active power loss of the system. i Let R be the current of the i-th line. i This is the resistance of the circuit.

[0027] Step 2 is implemented as follows:

[0028] Based on the basic data collected in Step 1, an initial population is generated, and preliminary calculations and evaluations are performed on the power loss and voltage curves of each individual. Each individual represents a distributed generation (DG) configuration. The purpose of this step is to provide an initial solution set for optimization by the quantum genetic algorithm.

[0029] 1. A population is initialized using a random number generation algorithm. This population consists of multiple different distributed generation (DG) configuration schemes. Each configuration scheme is called an "individual," and its "genes" indicate whether DG is deployed on each bus. For example, setting a bus to 1 indicates that DG is deployed on that bus, and setting it to 0 indicates that DG is not deployed on that bus. Through this gene encoding method, each individual actually represents a specific DG location and capacity configuration scheme.

[0030] 2. Substitute each generated individual (i.e., a specific DG configuration scheme) into the power system model and perform power flow calculations. For each individual configuration, calculate the net apparent power demand S of the system. r This reflects the difference between load power demand and DG power generation output.

[0031] The formula is as follows:

[0032] S r =S load -S DG (6)

[0033] Among them, S loadIt is the total apparent power demand of the system's current load; S DG It is the total apparent power output of the DG under the current individual configuration.

[0034] From the calculated net apparent power S r Extract the active power component, representing the active power loss P under the current configuration. DG,loss :

[0035] P DG,loss =real(S r (7)

[0036] Among them, real(S) r ) is S r The active power component. This value represents the power loss under the current configuration and serves as the basis for subsequent fitness assessments.

[0037] During power flow calculation, the voltage level of each bus is obtained. Ensure that the voltage of each bus is between 0.95 and 1.05 pu. If any bus voltage exceeds this range, mark that configuration as non-compliant with voltage requirements.

[0038] 3. Calculate the initial fitness value for each individual, which will be used in the subsequent optimization process of the quantum genetic algorithm. Fitness is mainly based on the active power loss P. DG,loss The fitness value is evaluated based on the magnitude of the power loss. Configurations with lower power loss have higher fitness. If a configuration's voltage exceeds the specified range, its fitness value is set lower to reduce the probability of that configuration being included in the next generation.

[0039] The initial fitness value Fitness(ψ) can be defined as follows:

[0040]

[0041] In this way, individuals with lower losses in the initial population have higher fitness, and are therefore more likely to be selected in subsequent optimization processes.

[0042] Step 3 is implemented as follows:

[0043] First, define the objective function, with the goal of minimizing the total active power loss of the system. The expression for the objective function is:

[0044]

[0045] Among them, P t I represents the total active power loss of the system. i Let R be the current of the i-th bus. iLet be the resistance of the line, and n represent the total number of lines in the system. The fitness function is used to evaluate the quality of an individual; a higher fitness indicates that the individual has lower losses and is more in line with the optimization objective.

[0046] The power balance relationship in the system is as follows:

[0047]

[0048] Among them, P Gi P represents the total active power output of the system, measured in kilowatts (kW); DG P represents the total active power output of a distributed generation (DG) system, measured in kilowatts (kW). i P represents the power demand (kW) of the load in the system. loss This indicates the system power loss.

[0049] Quantum genetic algorithms combine the features of quantum computing and classical genetic algorithms, using quantum registers to encode individuals and simulating the evolutionary mechanism of classical genetic algorithms. Their core lies in iteratively optimizing the population through subroutines of quantum selection, crossover, and mutation, gradually approaching the optimal solution to the problem.

[0050] 1. Initialization: First, set the algorithm parameters, including the population size n and the number of qubits c per individual, where n is divisible by 4 and c is even. The algorithm represents each individual as a quantum state consisting of c qubits and encodes each individual in a separate quantum register. The fitness function uses the quantum Hamiltonian H. P This indicates that the fitness of each individual is its fitness in H. P The energy value, the optimization goal is to make the population gradually evolve to the lowest energy state (i.e., H). P (ground state).

[0051] 2. Main Loop of the Algorithm: The main loop of the algorithm contains the following steps: quantum selection, quantum crossover, and quantum mutation operations. This loop continues until a termination condition is met (e.g., reaching the maximum number of generations or the fitness value no longer changing significantly). The main program of the algorithm includes:

[0052] (1) Quantum Selection Subroutine: The quantum selection subroutine sorts the population by fitness value, retains individuals with higher fitness, and discards individuals with lower fitness. The specific implementation is as follows:

[0053] a) Define the ground state of the problem: The Hamiltonian H... P The eigenbase (i.e., the set of energy eigenstates) is defined as the problem's ground state, denoted as |u|. k >. Each state |u k > Satisfy H P |u k >=∈ k |uk >, where ∈ k Let be eigenvalues, arranged in ascending order, such that ∈ k ≤∈ k+1 .

[0054] b) Ranking Network: Individuals are ranked by fitness using a quantum ranking network. The ranking process involves quantum comparison operators O. CMP and quantum commutation operator C SWAP Multi-layered combination.

[0055] Comparison operator O CMP Used to determine the fitness of two adjacent registers. Its input consists of two register states |u k > i and |u k’ > i+1 and an auxiliary qubit |0> a And adjust the state of the auxiliary bits according to the fitness status:

[0056]

[0057] • Controlled commutation operator C SWAP The sorting is completed based on the contents of the auxiliary bit status exchange register:

[0058]

[0059] After multiple comparisons and exchanges, individuals with higher fitness are placed at the front, while individuals with lower fitness are placed in the back and removed, thus achieving population selection.

[0060] (2) Quantum crossover subroutine: The purpose of the quantum crossover subroutine is to generate new individuals through quantum cloning and bit swapping. This operation mimics gene recombination in classical genetic algorithms.

[0061] a) Low-order register initialization: First, the low-order register is set to the reference state ρ0 to receive cloning results from individuals with higher fitness. Specifically, the population state after quantum selection can be represented as:

[0062]

[0063] Where tr low This indicates that trace operations are performed on individuals with low fitness, while individuals with high fitness are retained.

[0064] b) Quantum cloning: Using a quantum cloning machine (QCM) to clone highly fit individuals, the following two cloning methods can be selected:

[0065] • Biologically simulated quantum observation cloning (BCQO): The reference state is set as |1>, and its cloning operator U(θ,ρ0) satisfies the following conditions:

[0066]

[0067] Where θ is the observation operator.

[0068] · Universal Quantum Cloning Machine (UQCM): This cloning method has consistent cloning fidelity, and its cloning operator is defined as:

[0069]

[0070] in Let be the projection operator, and d be the dimension of the Hilbert space.

[0071] c) Quantum Bit Swapping: During the crossover process, individual characteristics are recombined by swapping some qubits in the two lower-order registers. The swapping process is represented by the following formula:

[0072]

[0073] Where S 2i+j,2i+c+j The exchange operator is used to exchange specified qubits.

[0074] (3) Quantum mutation subroutine: The quantum mutation subroutine is used to introduce moderate perturbation, increase population diversity, and avoid getting trapped in local optima.

[0075] a) Define the mutation operation: Quantum mutation is achieved by applying a random Pauli gate (X,Y,Z) to each qubit of an individual. The mutation probability is set to p. m And define all possible combinations of variant gates U μ .

[0076] The mutation operator can be represented as:

[0077]

[0078] in Indicates a possible Pauli gate, p μ This represents the probability of each door.

[0079] (4) Termination Conditions and Updates: The quantum mutation subroutine is used to introduce moderate perturbations to increase population diversity and avoid getting trapped in local optima. At the end of each generation iteration, it is checked whether the termination condition has been met, such as the maximum number of generations G or the fitness no longer changing significantly. Once the termination condition is met, the algorithm stops and outputs the individual with the highest fitness in the population as the final solution, which is the current optimal DG configuration. This optimal solution is the DG location and capacity configuration that minimizes power loss and meets the voltage level requirements of the system.

[0080] Step 4 is implemented as follows:

[0081] To ensure the adaptability of the optimized distributed generation (DG) configuration under different load conditions, dynamic load change simulations are performed. Specifically, by simulating load increases and decreases, the configuration location and capacity of DG are further verified and optimized to address the challenges posed by load fluctuations in real-world power distribution systems.

[0082] 1. Load Variation Settings: In load variation simulation, the load of the power distribution system is increased and decreased by a certain percentage. Typically, the load variation range is selected as ±25%, ±50%, ±75%, and ±100%. These variation ranges can cover the load fluctuations that the system may face, providing a sufficient test environment for verifying the robustness of the DG configuration.

[0083] 2. Dynamic Optimization Process: For each load change condition, repeat the quantum genetic algorithm optimization process defined in step 3. That is: reinitialize the population, substitute the changed load data into the power system model; optimize the configuration location and capacity of the DG through quantum selection, quantum crossover, and quantum mutation operations, and calculate the power loss and voltage level under each configuration; evaluate the fitness of each generation of optimized population to ensure that the DG configuration scheme can still achieve power loss minimization and voltage curve stability under different load conditions.

[0084] 3. Adaptive Assessment and Configuration Adjustment: Under each load change condition, record the final optimized DG configuration location and capacity, and evaluate the system's power loss and voltage curve stability. Typically, increased load leads to increased voltage curve volatility, but adaptive DG configuration can maintain power loss at a low level and keep voltage fluctuations within acceptable limits. Conversely, when the load decreases, the voltage curve becomes more stable, and power loss is further reduced. Based on the adaptive assessment results under different load conditions, the optimal DG configuration can be fine-tuned to further improve its robustness.

[0085] 4. Data Recording and Comparative Analysis: Record the optimal configuration results and performance indicators (such as power loss, voltage fluctuation amplitude, etc.) under different load conditions, and compare them with the results under the initial load conditions. This analysis helps to summarize the performance of the DG configuration under load changes and provides a reference for further system optimization.

[0086] 5. Final Configuration Scheme Extraction: Based on the optimization results under various load conditions, the optimal DG configuration scheme that performs best under multiple load conditions is extracted, or the optimal schemes under different load conditions are combined to ensure that the final scheme has good load adaptability in actual operation. The final scheme will achieve dynamic minimization of power loss while maintaining the stability of the voltage curve under different load conditions, providing support for the efficient and stable operation of the power distribution system.

[0087] Step 5 shall be implemented as follows:

[0088] Based on the aforementioned optimization process and the results of load change simulation, the final distributed generation (DG) configuration scheme is determined. The goal of this step is to output an optimal DG configuration that minimizes power loss and stabilizes voltage under various load conditions, in order to meet the high-efficiency operation requirements of the actual power distribution system.

[0089] 1. First, extract the optimal DG configuration scheme under various load conditions from the results of steps 3 and 4, including the specific DG installation location (bus number) and capacity. By comparing the optimal schemes under different load variations, select the configuration scheme with the most stable performance under each load condition.

[0090] 2. Among the extracted optimal configurations, the combination of DG location and capacity that performs well under most load conditions is selected as the final DG configuration scheme. This final scheme not only meets the requirements for optimal power loss and voltage curve stability under initial load conditions, but also maintains good adaptability when the load increases or decreases.

[0091] The final DG configuration will include the following specific parameters:

[0092] (1) DG location: The specific bus location of the optimal configuration.

[0093] (2) DG capacity: The DG capacity in the optimal scheme, in kW.

[0094] (3) Power loss and voltage curve: Under optimal configuration, the power loss value and voltage curve of the system ensure that the voltage is between the specified upper and lower limits (0.95 to 1.05 pu).

[0095] 3. For different load variation conditions (e.g., ±25%, ±50%), output the optimal DG configuration recommendation under those conditions to allow for flexible adjustments during actual operation. For example, if the system load increases to a certain level, the DG configuration can be adjusted accordingly based on the optimal solution under this load condition, thereby minimizing power loss and ensuring voltage curve stability.

[0096] The beneficial effects of this invention are as follows: It proposes a distributed generation (DG) optimization configuration method based on a quantum genetic algorithm, achieving the minimization of power loss and the improvement of voltage stability in the distribution system. This method can significantly reduce system power loss, and the optimized DG configuration scheme improves overall energy utilization efficiency through the scientific selection of location and capacity. Simultaneously, this invention sets voltage constraints and dynamic load simulations during the optimization process to ensure that the voltage curve fluctuates within an allowable range, meeting the power quality requirements of users. Through load change simulation and adaptive verification, the final output configuration scheme exhibits high stability and adaptability under various load conditions, improving the system's operational reliability and flexibility under load fluctuations. Furthermore, by effectively reducing power losses in the distribution network, this invention helps reduce the demand for fossil fuels and greenhouse gas emissions, thereby achieving a more environmentally friendly power supply. Therefore, this invention provides an efficient, stable, and environmentally friendly distributed generation optimization scheme for distribution systems, providing technical support for the construction of a modern smart grid. Attached Figure Description

[0097] Figure 1 This is a schematic diagram of a 6-node radial power distribution system according to the present invention;

[0098] Figure 2 This is a flowchart of the overall process of a method for minimizing power loss in a power distribution system based on quantum genetic algorithm to optimize the location and capacity of distributed power sources according to the present invention.

[0099] Figure 3 This invention provides a wiring diagram for an IEEE-34 node standard example.

[0100] Figure 4 This is a diagram showing the system line loss and voltage level before optimization according to the present invention;

[0101] Figure 5 This is a diagram showing the line loss and voltage level results after DG grid-connection optimization according to the present invention; Detailed Implementation

[0102] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0103] This invention discloses a method for minimizing power loss in a power distribution system based on quantum genetic algorithm to optimize the location and capacity of distributed power sources, which is implemented according to the following steps:

[0104] Step 1 is implemented as follows: First, data reading and system initialization are performed. By reading the core data of the power distribution network, the system can accurately represent the network structure. Specifically, the data includes: bus number and voltage level, impedance value of each line (resistance R). i and reactance X i Bus load (active power P) i and no Q i This includes the rated power and capacity limitations of distributed generation (DG) sources at their installation locations. These data collectively construct the distribution network topology, ensuring accurate bus-to-line connections and providing a foundation for subsequent power flow calculations and DG optimization.

[0105] After completing data reading, the system begins initial power flow calculation, employing a forward-backward sweep method for analysis. This method consists of two phases:

[0106] (1) Backward Sweep: Starting from the farthest bus in the system, the current on each line is calculated progressively towards the power source. Initially, it is assumed that the voltage of each bus is equal to the source voltage. For the load current of each bus, the following formula can be used:

[0107]

[0108] Among them, I ldi P represents the load current of the i-th bus. i and Q i These represent the active and reactive power demands on the bus, V. i This refers to the voltage of the busbars. During the backward scan, the current values ​​in the lines are continuously updated based on the voltage values ​​from the previous iteration until the currents of all buses in the system are initially distributed.

[0109] (2) Forward Sweep: Starting from the power source point and moving outwards, the voltage of each bus is calculated and updated progressively. Given the line impedance Z... i and current I Li Under the given conditions, the voltage of each busbar is calculated using the following formula:

[0110] V i =V u -Z i I Li (2)

[0111] Among them, V i V is the voltage of the i-th bus. u Z is the voltage of the upstream bus. iFor line impedance,

[0112] I Li This represents the current in the line. By continuously advancing towards the far-end busbar of the system, the voltage distribution of each busbar can be obtained, completing a preliminary analysis of the voltage profile.

[0113] After completing the initial power flow calculation, the system also needs to verify the initial voltage and power distribution. During this process, it is checked whether the voltage of each bus is within the specified allowable range (e.g., 0.95 to 1.05 pu). The formulas for calculating the upper and lower voltage limits are as follows:

[0114]

[0115] Among them, V base It is the system's reference voltage.

[0116] Simultaneously calculate the total power loss of the system. The formula for calculating the total power loss of the system is as follows:

[0117]

[0118] Among them, P loss I represents the total active power loss of the system. i Let R be the current of the i-th line. i This is the resistance of the circuit.

[0119] by Figure 1 Taking the radial power distribution system shown as an example, this system can use BBIC, BCBV

[0120] It is described by the DLF matrix.

[0121] (1) The BBIC matrix (Branch-Branch Incidence Current Matrix) represents the relationship between the branch currents using Kirchhoff's Current Law (KCL). Figure 1 In a 6-node system, a BBIC matrix can be constructed:

[0122]

[0123] Right now:

[0124] [B]=[BBIC][I] (7)

[0125] Where [B] is the branch current matrix, and [I] is the bus injected current matrix.

[0126] (2) The BCBV (Branch-Current Bus-Voltage Matrix) is used to represent the relationship between branch current and bus voltage. Voltage changes can be represented using the BCBV matrix as follows:

[0127]

[0128] Right now:

[0129] [ΔV]=[BCBV][B] (9) where ΔV represents the voltage difference matrix.

[0130] (3) The DLF matrix (Distribution Load Flow Matrix) combines the BBIC and BCBV matrices to form the relationship between voltage and injected current, which simplifies load flow calculation:

[0131] [ΔV]=[BCBV][BBIC][I]=[DLF][I] (10)

[0132] The DLF matrix can directly convert the injected current into changes in bus voltage, providing an efficient method for voltage level assessment in power flow calculation and optimization processes.

[0133] Step 2 is implemented as follows: Based on the basic data collected in Step 1, an initial population is generated, and the power loss and voltage curve of each individual are preliminarily calculated and evaluated.

[0134] Each individual represents a distributed generation (DG) configuration. The purpose of this step is to provide an initial solution set for optimization by the quantum genetic algorithm.

[0135] 1. A population is initialized using a random number generation algorithm. This population consists of multiple different distributed generation (DG) configuration schemes. Each configuration scheme is called an "individual," and its "genes" indicate whether DG is deployed on each bus. For example, setting a bus to 1 indicates that DG is deployed on that bus, and setting it to 0 indicates that DG is not deployed on that bus. Through this gene encoding method, each individual actually represents a specific DG location and capacity configuration scheme.

[0136] 2. Substitute each generated individual (i.e., a specific DG configuration scheme) into the power system model and perform power flow calculations. For each individual configuration, calculate the net apparent power demand S of the system. r This reflects the difference between load power demand and DG power generation output.

[0137] The formula is as follows:

[0138] S r =S load-S DG (11)

[0139] Among them, S load It is the total apparent power demand of the system's current load; S DG It is the total apparent power output of the DG under the current individual configuration.

[0140] From the calculated net apparent power S r Extract the active power component, representing the active power loss P under the current configuration. DG,loss :

[0141] P DG,loss =real(S r (12)

[0142] Among them, real(S) r ) is S r The active power component. This value represents the power loss under the current configuration and serves as the basis for subsequent fitness assessments.

[0143] During power flow calculation, the voltage level of each bus is obtained. Ensure that the voltage of each bus is between 0.95 and 1.05 pu. If any bus voltage exceeds this range, mark that configuration as non-compliant with voltage requirements.

[0144] 3. Calculate the initial fitness value for each individual, which will be used in the subsequent optimization process of the quantum genetic algorithm. Fitness is mainly based on the active power loss P. DG,loss The fitness value is evaluated based on the magnitude of the power loss. Configurations with lower power loss have higher fitness. If a configuration's voltage exceeds the specified range, its fitness value is set lower to reduce the probability of that configuration being included in the next generation.

[0145] The initial fitness value Fitness(ψ) can be defined as follows:

[0146]

[0147] In this way, individuals with lower losses in the initial population have higher fitness, and are therefore more likely to be selected in subsequent optimization processes.

[0148] Step 3 is implemented as follows: First, define the objective function, with the optimization objective being to minimize the total active power loss of the system. The expression for the objective function is:

[0149]

[0150] Among them, P t I represents the total active power loss of the system. i Let R be the current of the i-th bus. iLet be the resistance of the line, and n represent the total number of lines in the system. The fitness function is used to evaluate the quality of an individual; a higher fitness indicates that the individual has lower losses and is more in line with the optimization objective.

[0151] The power balance relationship in the system is as follows:

[0152]

[0153] Among them, P Gi P represents the total active power output of the system, measured in kilowatts (kW); DG P represents the total active power output of a distributed generation (DG) system, measured in kilowatts (kW). i P represents the power demand (kW) of the load in the system. loss This indicates the system power loss.

[0154] Quantum genetic algorithms combine the features of quantum computing and classical genetic algorithms, using quantum registers to encode individuals and simulating the evolutionary mechanism of classical genetic algorithms. Their core lies in iteratively optimizing the population through subroutines of quantum selection, crossover, and mutation, gradually approaching the optimal solution to the problem.

[0155] 1. Initialization: First, set the algorithm parameters, including the population size n and the number of qubits c per individual, where n is divisible by 4 and c is even. The algorithm represents each individual as a quantum state consisting of c qubits and encodes each individual in a separate quantum register. The fitness function uses the quantum Hamiltonian H. P This indicates that the fitness of each individual is its fitness in H. P The energy value, the optimization goal is to make the population gradually evolve to the lowest energy state (i.e., H). P (ground state).

[0156] 2. Main Loop of the Algorithm: The main loop of the algorithm contains the following steps: quantum selection, quantum crossover, and quantum mutation operations. This loop continues until a termination condition is met (e.g., reaching the maximum number of generations or the fitness value no longer changing significantly). The main program of the algorithm includes:

[0157] (1) Quantum Selection Subroutine: The quantum selection subroutine sorts the population by fitness value, retains individuals with higher fitness, and discards individuals with lower fitness. The specific implementation is as follows:

[0158] a) Define the ground state of the problem: The Hamiltonian H... P The eigenbase (i.e., the set of energy eigenstates) is defined as the problem's ground state, denoted as |u|. k >. Each state |u k > Satisfy H P |u k >=∈ k |uk >, where ∈ k Let be eigenvalues, arranged in ascending order, such that ∈ k ≤∈ k+1 .

[0159] b) Ranking Network: Individuals are ranked by fitness using a quantum ranking network. The ranking process involves quantum comparison operators O. CMP and quantum commutation operator C SWAP Multi-layered combination.

[0160] Comparison operator O CMP Used to determine the fitness of two adjacent registers. Its input consists of two register states |u k > i and |u k’ > i+1 and an auxiliary qubit |0> a And adjust the state of the auxiliary bits according to the fitness status:

[0161]

[0162] • Controlled commutation operator C SWAP The sorting is completed based on the contents of the auxiliary bit status exchange register:

[0163]

[0164] After multiple comparisons and exchanges, individuals with higher fitness are placed at the front, while individuals with lower fitness are placed in the back and removed, thus achieving population selection.

[0165] (2) Quantum crossover subroutine: The purpose of the quantum crossover subroutine is to generate new individuals through quantum cloning and bit swapping. This operation mimics gene recombination in classical genetic algorithms.

[0166] a) Low-order register initialization: First, the low-order register is set to the reference state ρ0 to receive cloning results from individuals with higher fitness. Specifically, the population state after quantum selection can be represented as:

[0167]

[0168] Where tr low This indicates that trace operations are performed on individuals with low fitness, while individuals with high fitness are retained.

[0169] b) Quantum cloning: Using a quantum cloning machine (QCM) to clone highly fit individuals, the following two cloning methods can be selected:

[0170] • Biologically simulated quantum observation cloning (BCQO): The reference state is set as |1>, and its cloning operator U(θ,ρ0) satisfies the following conditions:

[0171]

[0172] Where θ is the observation operator.

[0173] · Universal Quantum Cloning Machine (UQCM): This cloning method has consistent cloning fidelity, and its cloning operator is defined as:

[0174]

[0175] in Let be the projection operator, and d be the dimension of the Hilbert space.

[0176] c) Quantum Bit Swapping: During the crossover process, individual characteristics are recombined by swapping some qubits in the two lower-order registers. The swapping process is represented by the following formula:

[0177]

[0178] Where S 2i+j,2i+c+j The exchange operator is used to exchange specified qubits.

[0179] (3) Quantum mutation subroutine: The quantum mutation subroutine is used to introduce moderate perturbation, increase population diversity, and avoid getting trapped in local optima.

[0180] a) Define the mutation operation: Quantum mutation is achieved by applying a random Pauli gate (X,Y,Z) to each qubit of an individual. The mutation probability is set to p. m And define all possible combinations of variant gates U μ .

[0181] The mutation operator can be represented as:

[0182]

[0183] in Indicates a possible Pauli gate, p μ This represents the probability of each door.

[0184] (4) Termination Conditions and Updates: The quantum mutation subroutine is used to introduce moderate perturbations to increase population diversity and avoid getting trapped in local optima. At the end of each generation iteration, it is checked whether the termination condition has been met, such as the maximum number of generations G or the fitness no longer changing significantly. Once the termination condition is met, the algorithm stops and outputs the individual with the highest fitness in the population as the final solution, which is the current optimal DG configuration. This optimal solution is the DG location and capacity configuration that minimizes power loss and meets the voltage level requirements of the system.

[0185] Step 4 is implemented as follows: To ensure the adaptability of the optimized distributed generation (DG) configuration under different load conditions, dynamic load change simulation is performed. Specifically, by simulating load increases and decreases, the configuration location and capacity of DG are further verified and optimized to address the challenges posed by load fluctuations in actual power distribution systems.

[0186] 1. Load Variation Settings: In load variation simulation, the load of the power distribution system is increased and decreased by a certain percentage. Typically, the load variation range is selected as ±25%, ±50%, ±75%, and ±100%. These variation ranges can cover the load fluctuations that the system may face, providing a sufficient test environment for verifying the robustness of the DG configuration.

[0187] 2. Dynamic Optimization Process: For each load change condition, repeat the quantum genetic algorithm optimization process defined in step 3. That is: reinitialize the population, substitute the changed load data into the power system model; optimize the configuration location and capacity of the DG through selection, crossover, and mutation operations, and calculate the power loss and voltage level under each configuration; evaluate the fitness of each generation of optimized population to ensure that the DG configuration scheme can still achieve power loss minimization and voltage curve stability under different load conditions.

[0188] 3. Adaptive Assessment and Configuration Adjustment: Under each load change condition, record the final optimized DG configuration location and capacity, and evaluate the system's power loss and voltage curve stability. Typically, increased load leads to increased voltage curve volatility, but adaptive DG configuration can maintain power loss at a low level and keep voltage fluctuations within acceptable limits. Conversely, when the load decreases, the voltage curve becomes more stable, and power loss is further reduced. Based on the adaptive assessment results under different load conditions, the optimal DG configuration can be fine-tuned to further improve its robustness.

[0189] 4. Data Recording and Comparative Analysis: Record the optimal configuration results and performance indicators (such as power loss, voltage fluctuation amplitude, etc.) under different load conditions, and compare them with the results under the initial load conditions. This analysis helps to summarize the performance of the DG configuration under load changes and provides a reference for further system optimization.

[0190] 5. Final Configuration Scheme Extraction: Based on the optimization results under various load conditions, the optimal DG configuration scheme that performs best under multiple load conditions is extracted, or the optimal schemes under different load conditions are combined to ensure that the final scheme has good load adaptability in actual operation. The final scheme will achieve dynamic minimization of power loss while maintaining the stability of the voltage curve under different load conditions, providing support for the efficient and stable operation of the power distribution system.

[0191] Step 5 is implemented as follows: Based on the aforementioned optimization process and the results of load change simulation, the final distributed generation (DG) configuration scheme is determined. The goal of this step is to output an optimal DG configuration that minimizes power loss and stabilizes voltage under various load conditions, in order to meet the high-efficiency operation requirements of the actual power distribution system.

[0192] 1. First, extract the optimal DG configuration scheme under various load conditions from the results of steps 3 and 4, including the specific DG installation location (bus number) and capacity. By comparing the optimal schemes under different load variations, select the configuration scheme with the most stable performance under each load condition.

[0193] 2. Among the extracted optimal configurations, the combination of DG location and capacity that performs well under most load conditions is selected as the final DG configuration scheme. This final scheme not only meets the requirements for optimal power loss and voltage curve stability under initial load conditions, but also maintains good adaptability when the load increases or decreases.

[0194] The final DG configuration will include the following specific parameters:

[0195] • DG location: The specific bus location for the optimal configuration.

[0196] • DG capacity: The DG capacity in the optimal scheme, in kW.

[0197] • Power loss and voltage curve: Under optimal configuration, the system's power loss value and voltage curve ensure that the voltage is between the specified upper and lower limits (0.95 to 1.05 pu).

[0198] 3. For different load variation conditions (e.g., ±25%, ±50%), output the optimal DG configuration recommendation under those conditions to allow for flexible adjustments during actual operation. For example, if the system load increases to a certain level, the DG configuration can be adjusted accordingly based on the optimal solution under this load condition, thereby minimizing power loss and ensuring voltage curve stability.

[0199] Overall flowchart as follows Figure 2 As shown.

[0200] Example

[0201] This invention is based on the IEEE-34-node system, and related simulation verification is performed. Its single-line diagram and wiring method are as follows: Figure 3 As shown in the figure. The network rated voltage is 24.9kV, and the active and reactive loads are 1769kW and 1044kVAR, respectively. Considering the diverse nonlinear loads of the distribution network, three independent harmonic sources—electric vehicle charging stations (EVC), adjustable speed drives (ASD), and variable-frequency drives (VFD)—are connected to simulate the diverse nonlinear loads. For ease of comparison and analysis, the system parameters of the example before optimization are first introduced, and the specific results are as follows. Figure 4 As shown in the figure, the left axis represents active power loss, corresponding to the bar chart below, which shows the active power loss of each line in the system. The table below also lists the total line loss and line loss rate of the system. The right axis represents the voltage level of the system before optimization, and the table above lists the maximum positive deviation, maximum negative deviation, and mean deviation of the system voltage at this time.

[0202] The optimized scheme for grid connection is shown in Table 1, where the DG grid connection point and capacity are 854 kW and 532.656 kW, respectively. The distribution of line losses and voltage in the optimized system is shown in Table 1. Figure 5 As shown.

[0203] Table 1 Optimization Plan for Grid Connection Points

[0204]

[0205] Depend on Figure 5 It can be seen that after adopting the optimized access scheme, the system bus loss is 191.884kW, and the line loss rate is 9.5%. At this time, the main line losses are concentrated on L3, L5, L6, L14, L27, and L32. As for voltage distribution, the maximum positive and negative deviations and the average voltage deviation are 0.049pu, -0.049pu, and 0.042pu, respectively.

Claims

1. A method for minimizing power loss in a power distribution system based on quantum genetic algorithm to optimize the location and capacity of distributed power sources, characterized in that, The specific steps are as follows: Step 1: First, collect basic data of the power distribution network, including specific information on buses, lines, and distributed generation equipment, such as capacity and location, and initialize the voltage level and power flow of the power system to obtain the power loss and voltage curve of the system in the initial state. Step 2: After obtaining the basic data, a random number generation algorithm is used to generate an initial population. This initial population contains several different distributed generation (DG) configuration schemes. Each configuration scheme is called an "individual". The individual is composed of genes, and each gene corresponds to a bus, which is used to indicate whether DG is deployed on the bus. These individuals are substituted into the initial power system model to calculate their corresponding power loss and voltage curves, so as to provide a preliminary evaluation for subsequent optimization. Step 3: After obtaining the initial population, a quantum genetic algorithm is used to iteratively optimize the population, aiming to minimize power loss and keep the voltage curve within the allowable range. The optimization process of the quantum genetic algorithm includes quantum selection, quantum crossover, and quantum mutation. First, quantum selection is based on the fitness function, selecting individuals with low power loss and stable voltage to enter the next generation. Next, quantum crossover uses quantum superposition and cloning technology to combine the characteristics of high-fitness individuals to generate new individuals, thereby gradually optimizing the DG configuration. Finally, quantum mutation introduces perturbation through random Pauli gates to increase population diversity and avoid local optima. In each generation, a new generation of configuration is generated through quantum selection, crossover, and mutation, gradually obtaining a better DG location and capacity scheme. Step 4: After obtaining the preliminary optimized configuration, simulate different load conditions of the power distribution system to verify the adaptability of the DG configuration; specifically, set the load conditions to increase or decrease by a certain percentage, such as 25%, 50% or other preset percentages, and repeat the above quantum genetic algorithm optimization steps to ensure that the configuration scheme can still achieve minimum power loss and voltage stability under different load conditions; record the optimal configuration under each load condition, and evaluate the changes in the power loss and voltage curve of the system under each load condition; Step 5: Finally, output the optimal DG location and capacity configuration scheme, including the DG location and capacity of each bus, and the optimized system power loss and voltage curve data; at the same time, provide corresponding optimized configurations for different load conditions to ensure that the DG configuration can adapt to the dynamic changes of the load, thereby achieving stable and efficient power supply capability of the system.

2. The method for minimizing power loss in a power distribution system based on optimizing the location and capacity of distributed power sources using a quantum genetic algorithm, as described in claim 1, is characterized in that... Step 1 is implemented as follows: First, data reading and system initialization are performed. By reading the core data of the power distribution network, the system can accurately represent the network structure. Specifically, the data includes: busbar numbers and voltage levels; the resistance R of each line. i and reactance X i Bus load, including active power P i and reactive power Q i ; and the rated power and capacity limitations of the DG at the installation location; After completing the data reading, the system begins the initial power flow calculation, employing the forward-backward sweep method for analysis; this method consists of two stages: (1) Backward Sweep: Starting from the farthest bus in the system, calculate the current on each line step by step towards the power source; initially, assume that the voltage of each bus is equal to the source voltage; for the load current of each bus, the following formula can be used: Among them, I ldi P represents the load current of the i-th bus. i and Q i These represent the active and reactive power demands on the bus, V. i It is the voltage of the bus; during the backward scan, the current value in the line will be continuously updated according to the voltage value of the previous iteration until the current of all buses in the system is initially distributed. (2) Forward Sweep: Starting from the power source point and moving outwards, the voltage of each bus is calculated and updated progressively; given the line impedance Z... i and current I Li Under the given conditions, the voltage of each busbar is calculated using the following formula: V i =V u -Z i I Li (2) Among them, V i V is the voltage of the i-th bus. u Z is the voltage of the upstream bus. i I is the line impedance. Li The current in the line; by continuously advancing towards the far-end busbar of the system, the voltage distribution of each busbar can be obtained, and a preliminary analysis of the voltage profile can be completed. After completing the initial power flow calculation, the system also needs to verify the initial voltage and power distribution. During this process, it checks whether the voltage of each bus is within the specified allowable range, which is 0.95 pu to 1.05 pu. The calculation formula for the upper and lower voltage limits is as follows: Among them, V base It is the system's reference voltage; Simultaneously calculate the total power loss of the system; the formula for calculating the total power loss of the system is as follows: Among them, P loss I represents the total active power loss of the system. i Let R be the current of the i-th line. i This is the resistance of the circuit.

3. The method for minimizing power loss in a power distribution system based on optimizing the location and capacity of distributed power sources using a quantum genetic algorithm, as described in claim 1, is characterized in that... Step 2 is implemented as follows: Based on the basic data collected in step 1, an initial population is generated, and the power loss and voltage curve of each individual are preliminarily calculated and evaluated; each individual represents a DG configuration scheme; the purpose of this step is to provide an initial solution set for the optimization of the quantum genetic algorithm. Step 2.1: Initialize a population using a random number generation algorithm. These populations consist of multiple DG configuration schemes. Each configuration scheme is called an "individual". The "gene" in the individual indicates whether a DG is placed on each bus. Setting a bus to 1 indicates that a DG is placed on the bus, and setting it to 0 indicates that a DG is not placed on the bus. Through this gene encoding method, each individual actually represents a specific DG location and capacity configuration scheme. Step 2.2: Substitute each generated individual into the power system model. Each individual corresponds to a specific DG configuration scheme, and perform power flow calculations on it. For each individual configuration, calculate the net apparent power demand S of the system. r This reflects the difference between load power demand and DG power generation output; the formula is as follows: S r =S load -S DG (6) Among them, S load It is the total apparent power demand of the system's current load; S DG This is the total apparent power output of the DG under the current individual configuration; From the calculated net apparent power S r Extract the active power component, representing the active power loss P under the current configuration. DG,loss : P DG,loss =real(S r (7) Where, real(S) r ) is S r The active power component; this value represents the power loss under the current configuration and serves as the basis for subsequent fitness assessments; During the power flow calculation, the voltage level of each bus is obtained; ensure that the voltage of each bus is between 0.95 and 1.05 pu. If the voltage of some buses exceeds this range, mark the individual configuration as a configuration that does not meet the voltage requirements. Step 2.3: Calculate the initial fitness value for each individual, which will be used in the subsequent optimization process of the quantum genetic algorithm; the fitness is mainly based on the active power loss P. DG,loss The fitness value is evaluated based on the magnitude of the power loss; configurations with lower power loss have higher fitness; if the voltage of a configuration exceeds the specified range, its fitness value is set lower to reduce the probability of that configuration being included in the next generation. The initial fitness value Fitness(ψ) can be defined as follows:

4. The method for minimizing power loss in a power distribution system based on optimizing the location and capacity of distributed power sources using a quantum genetic algorithm, as described in claim 1, is characterized in that... Step 3 is implemented as follows: First, define the objective function, with the goal of minimizing the total active power loss of the system; the expression of the objective function is: Among them, P t I represents the total active power loss of the system. i Let R be the current of the i-th bus. i Let n be the resistance of the line, and n represent the total number of lines in the system. The power balance relationship in the system is as follows: Among them, P Gi P represents the total active power output of the system, measured in kilowatts (kW); DG P represents the total active power output of the DG system, measured in kilowatts (kW); i P represents the power demand (kW) of the load in the system. loss Indicates system power loss; Step 3.1 Initialization: First, set the algorithm parameters, including the population size n and the number of qubits c for each individual, where n is divisible by 4 and c is even. The algorithm represents each individual as a quantum state consisting of c qubits and encodes each individual in a separate quantum register. The fitness function uses the quantum Hamiltonian H. P This indicates that the fitness of each individual is its fitness in H. P The energy value is determined by optimizing the population to gradually evolve to the lowest energy state, which is H. P The ground state; Step 3.2, Main Loop of the Algorithm: The main loop of the algorithm includes the following steps: quantum selection, quantum crossover, and quantum mutation operations; this loop continues until the termination condition is met; in the main program of the algorithm, it includes: Step 3.2.1, Quantum Selection Subroutine: The quantum selection subroutine sorts the population by fitness value, retains individuals with higher fitness, and discards individuals with lower fitness; the specific implementation is as follows: Step 3.2.1.1, Define the ground state of the problem: Define the quantum Hamiltonian H. P The eigenbase is defined as the ground state of the problem, and the eigenbase is the set of energy eigenstates, denoted as |u|. k >;Each state |u k > Satisfies H P |u k >=∈ k |u k >, where ∈ k Let be eigenvalues, arranged in ascending order, such that ∈ k ≤∈ k+1 ; Step 3.2.1.2, Sorting Network: Individuals are sorted by fitness using a quantum sorting network; the sorting process includes quantum comparison operators O. CMP and quantum commutation operator C SWAP Multi-layered combination; Comparison operator O CMP Used to determine the fitness of two adjacent registers; its input is two register states |u k < i and |u k '> i+1 and an auxiliary qubit |0> a And adjust the state of the auxiliary bits according to the fitness status: Controlled commutator C SWAP The sorting is completed based on the contents of the auxiliary bit status exchange register: After multiple comparisons and exchanges, individuals with higher fitness were placed at the front, while individuals with lower fitness were placed in the back and removed. Step 3.2.2, Quantum Crossover Subroutine: Step 3.2.2.1, Low-order register initialization: First, set the low-order register to the reference state ρ0; specifically, the population state after quantum selection can be represented as: Where tr low This indicates that trace operations are performed on individuals with low fitness, while individuals with high fitness are retained; symbol Represents tensor product operation; Step 3.2.2.2, Quantum Cloning: A quantum cloning machine (QCM) is used to clone individuals with high fitness. The following two cloning methods can be selected: (1) Biologically simulated quantum observation cloning (BCQO): The reference state is set as |1>, and its cloning operator U(θ,ρ0) satisfies the following conditions: Where θ is the observation operator; symbol This indicates Hermitian conjugate. (2) Universal Quantum Cloning Machine (UQCM): This cloning method has consistent cloning fidelity, and its cloning operator is defined as: in Let be the projection operator, and d be the dimension of the Hilbert space; Step 3.2.2.3, Quantum Bit Swapping: During the crossover process, individual characteristics are recombined by swapping some qubits in the two lower-order registers; the swapping process is represented by the following formula: Where S 2i+j,2i+c+j The exchange operator is used to exchange specified qubits; Step 3.2.3, Quantum Mutation Subroutine: Quantum mutation occurs through each qubit of an individual with probability p. m This is achieved by applying a random Pauli gate (X,Y,Z); if mutation occurs, the specific Pauli gate for the i-th bit is determined by the probability distribution {p μ } determines; thereby defining all possible combinations of variant gates U μ Its operator representation is: Where n represents the number of qubits contained in an individual; Indicates a possible Pauli gate; μ i For the corresponding operation label, This indicates the selection gate on the i-th bit. The probability of and satisfying Step 3.2.4, Termination Condition and Update: At the end of each generation iteration, check whether the termination condition has been met. The termination condition is the maximum number of generations G or the fitness no longer changes significantly. Once the termination condition is met, the algorithm stops and outputs the individual with the highest fitness in the population as the final solution, which is the current optimal DG configuration scheme. This optimal scheme is the DG location and capacity configuration with the minimum power loss and the voltage level that meets the system requirements.

5. The method for minimizing power loss in a power distribution system based on optimizing the location and capacity of distributed power sources using a quantum genetic algorithm, as described in claim 1, is characterized in that... Step 4 is implemented as follows: Step 4.1, Load Change Settings: In the load change simulation, the load of the power distribution system is increased and decreased by a certain percentage. The selected load change ranges are ±25%, ±50%, ±75%, and ±100%. Step 4.2, Dynamic Optimization Process: For each load change condition, repeat the quantum genetic algorithm optimization process defined in Step 3; That is: reinitialize the population, substitute the changed load data into the power system model; optimize the configuration location and capacity of the DG through selection, crossover and mutation operations, and calculate the power loss and voltage level under each configuration; Step 4.3 Adaptability Assessment and Configuration Adjustment: Under each load change condition, record the final optimized DG configuration location and capacity, and evaluate the system's power loss and voltage curve stability; Step 4.4, Data Recording and Comparative Analysis: Record the optimal configuration results and performance indicators under different load conditions, including power loss and voltage fluctuation amplitude, and compare and analyze them with the results under the initial load conditions. Step 4.5: Final configuration scheme extraction: Based on the optimization results under various load conditions, extract the DG configuration scheme that performs best under multiple load conditions, or combine the best schemes under different load conditions to ensure that the final scheme has good load adaptability in actual operation.

6. The method for minimizing power loss in a power distribution system based on optimizing the location and capacity of distributed power sources using a quantum genetic algorithm, as described in claim 1, is characterized in that... Step 5 is implemented as follows: Based on the aforementioned optimization process and the results of load change simulation, the final distributed generation configuration scheme is determined. Step 5.1 First, extract the optimal DG configuration scheme under various load conditions from the results of Step 3 and Step 4, including the specific DG installation location and corresponding bus number, as well as the capacity. Step 5.2: From the extracted optimal configurations, select the DG location and capacity combination that performs well under most load conditions as the final DG configuration scheme. The final DG configuration will include the following specific parameters: (1) DG location: The specific bus location of the optimal configuration; (2) DG capacity: The DG capacity in the optimal scheme, in kW; (3) Power loss and voltage curve: Under the optimal configuration, the power loss value and voltage curve of the system ensure that the voltage is between the specified upper and lower limits, which are 0.95pu to 1.05pu; Step 5.3: For different load change conditions, output the DG configuration recommendation that performs best under those conditions.

Images