CN105606931A

CN105606931A - Quantum-genetic-algorithm-based fault diagnosis method for medium-voltage distribution network

Info

Publication number: CN105606931A
Application number: CN201511030548.1A
Authority: CN
Inventors: 姚瑛; 郗晓光; 王浩鸣; 李琳; 董艳唯
Original assignee: State Grid Corp of China SGCC; State Grid Tianjin Electric Power Co Ltd
Current assignee: State Grid Corp of China SGCC; State Grid Tianjin Electric Power Co Ltd
Priority date: 2015-12-30
Filing date: 2015-12-30
Publication date: 2016-05-25

Abstract

The present invention relates to a fault diagnosis method for medium-voltage distribution network based on quantum genetic algorithm. Its technical features include the following steps: Step 1. Using the actual and expected values of component actions and integrating them into circuit breaker failure protection and circuit breaker automatic reclosing Establish an improved medium-voltage distribution network fault diagnosis model based on the protection state; step 2, solve the improved medium-voltage distribution network fault diagnosis model based on the quantum genetic algorithm, and perform fault diagnosis on the medium-voltage distribution network. The invention establishes a distribution network fault diagnosis model by analyzing the types of components in the power distribution system and the protection configuration, simulates the protection and the action of the circuit breaker after the fault scene, and uses the quantum genetic algorithm to solve the fault model, thereby realizing the fault diagnosis in the power distribution system. After the distribution network fails, it can accurately locate the fault component and quickly and comprehensively diagnose the distribution network fault.

Description

A Fault Diagnosis Method for Medium Voltage Distribution Network Based on Quantum Genetic Algorithm

技术领域technical field

本发明涉及城市配电网规划以及故障诊断技术领域，特别是一种基于量子遗传算法的中压配电网故障诊断方法。The invention relates to the technical field of urban distribution network planning and fault diagnosis, in particular to a method for fault diagnosis of a medium-voltage distribution network based on a quantum genetic algorithm.

背景技术Background technique

配电系统作为电能生产、传输和使用的重要环节，是联系实际用户需求侧与发、输电系统的关键纽带。因此，如何在配电网发生故障后对故障进行有效、合理的定位并加以诊断是当前城市电网评估发展的关键部分。考虑到在配电网中，通信信息装置所处的恶劣环境以及不同地区配电自动化发展程度的不一致等原因造成了配电网的故障信息中存在着大量不确定因素；并且一旦在配电网中发生多重复杂故障，在失电区域中将会存在故障元件与非故障元件，考虑到配电系统规模的庞大，元件数量与种类繁多，很难在短时间内确定出元件是否处于故障状态，加之配电网中的线路保护装置或断路器会发生拒动或误动的情况，因此，上述这些因素都会导致配电网故障分析的范围扩大、故障信息在上传相关部门的过程中发生畸变等而导致无法准确确定故障元件，给配电系统的安全稳定运行带来了危害。而随着配电系统接线形式的日趋复杂、设备元件不断增多、设备规模不断增大，并且用户需求侧对于供电的要求逐步提高。因此，有效、合理的对配电系统进行故障诊断对于电力系统的综合发展以及需求侧的可靠用电均具有十分重要的意义。As an important link in the production, transmission and use of electric energy, the power distribution system is the key link between the actual user demand side and the power generation and transmission system. Therefore, how to effectively and reasonably locate and diagnose the fault after the fault occurs in the distribution network is a key part of the current evaluation and development of the urban power grid. Considering that in the distribution network, the harsh environment of the communication information device and the inconsistency in the development of distribution automation in different regions have caused a large number of uncertain factors in the fault information of the distribution network; and once in the distribution network Multiple complex faults occur in the power failure area. There will be faulty components and non-faulty components in the power-off area. Considering the large scale of the power distribution system, the number and variety of components, it is difficult to determine whether the components are in a faulty state in a short time. In addition, the line protection devices or circuit breakers in the distribution network may refuse to operate or malfunction. Therefore, the above factors will lead to the expansion of the scope of distribution network fault analysis, and the distortion of fault information in the process of uploading to relevant departments, etc. As a result, the failure components cannot be accurately determined, which brings harm to the safe and stable operation of the power distribution system. With the increasingly complex wiring forms of the power distribution system, the increasing number of equipment components, and the increasing scale of equipment, the requirements for power supply on the user demand side are gradually increasing. Therefore, effective and reasonable fault diagnosis of power distribution system is of great significance to the comprehensive development of power system and reliable power consumption on the demand side.

目前，针对电力系统的故障诊断方法较多，主要思路都是通过电力系统中开关元件的动作信息进行故障判断与分析。目前有关电力系统故障诊断的方法主要有：(1)基于粗糙集与决策树的配电网故障诊断算法，主要是利用了粗糙集具有较好处理不确定信息的能力，实现了对故障样本决策表的故障规则自取；(2)基于概率神经网络(PNN)的高压断路器故障诊断方法，有效分析高压断路器的故障特性，进行故障定位；(3)粒子群算法与神经网络相结合的模拟电路故障诊断方法，将故障信号进行有效分解，再通过归一化处理提取故障特征信息并以此做为神经网络的输入学习样本；(4)基于时序模糊Petri网的故障诊断方法，通过建立故障诊断模型，完成对继电保护动作的评价。以上方法为电力系统的故障诊断提供了良好的研究思路，但仍存在以下局限：第一，未能考虑配电网故障情况下保护或开关拒动、误动以及信息畸变时的诊断准确性；第二，电力系统在故障情况下元件动作的时序性没有被充分考虑。At present, there are many fault diagnosis methods for power systems, and the main idea is to judge and analyze faults through the action information of switching elements in the power system. At present, the methods of power system fault diagnosis mainly include: (1) distribution network fault diagnosis algorithm based on rough set and decision tree, which mainly utilizes the ability of rough set to deal with uncertain information better, and realizes the decision-making of fault samples (2) The high-voltage circuit breaker fault diagnosis method based on the probabilistic neural network (PNN), which can effectively analyze the fault characteristics of the high-voltage circuit breaker and perform fault location; (3) The combination of particle swarm optimization and neural network The analog circuit fault diagnosis method effectively decomposes the fault signal, and then extracts the fault feature information through normalization processing and uses it as the input learning sample of the neural network; (4) The fault diagnosis method based on time series fuzzy Petri net The fault diagnosis model completes the evaluation of the relay protection action. The above method provides a good research idea for the fault diagnosis of power system, but there are still the following limitations: First, it fails to consider the diagnostic accuracy of protection or switch refusal, malfunction and information distortion in the case of distribution network faults; Second, the timing of component actions in the power system under fault conditions has not been fully considered.

相较于上述几种方法，量子遗传算法(QuantumGeneticAlgorithm,QGA)将量子理论有效地融入到经典遗传算法当中，比传统遗传算法具有搜索范围更广，全局寻优的搜索效率更高，适应性更强等优势，并且能够保证算法的收敛性。Compared with the above methods, the Quantum Genetic Algorithm (QGA) effectively integrates the quantum theory into the classical genetic algorithm, and has a wider search range than the traditional genetic algorithm. Strong and other advantages, and can guarantee the convergence of the algorithm.

发明内容Contents of the invention

本发明的目的在于克服现有技术的不足，提供一种设计合理、故障诊断全面、分析准确且定位快速的基于量子遗传算法的中压配电网故障诊断方法。The purpose of the present invention is to overcome the deficiencies of the prior art, and provide a medium-voltage distribution network fault diagnosis method based on quantum genetic algorithm, which is reasonable in design, comprehensive in fault diagnosis, accurate in analysis and fast in positioning.

本发明解决其技术问题是采取以下技术方案实现的：The present invention solves its technical problem and realizes by taking the following technical solutions:

一种基于量子遗传算法的中压配电网故障诊断方法，包括以下步骤：A method for fault diagnosis of medium-voltage distribution network based on quantum genetic algorithm, comprising the following steps:

步骤1、采用元件动作的实际与期望值并融入断路器失灵保护以及断路器自动重合闸的保护状态建立改进型中压配电网故障诊断模型；该模型的目标函数为：Step 1. Establish an improved medium-voltage distribution network fault diagnosis model by using the actual and expected values of component actions and incorporating the protection status of circuit breaker failure protection and automatic reclosing of circuit breakers; the objective function of the model is:

$\begin{matrix} E E. ((x x)) = = {Σ Σ}_{k k = = 11}^{n no} | | {r r}_{k k,, m m} - - {r r}_{k k,, m m}^{* *} | | | | 11 - - {r r}_{k k,, s the s} {r r}_{k k,, s the s}^{* *} - - \overset{&CirclePlus; &CirclePlus;}{Σ Σ} {r r}_{k k,, l l} {r r}_{k k,, l l}^{* *} | | + + {Σ Σ}_{k k = = 11}^{n no} | | {r r}_{k k,, s the s} - - {r r}_{k k,, s the s}^{* *} | | | | 11 - - \overset{&CirclePlus; &CirclePlus;}{Σ Σ} {r r}_{k k,, l l} {r r}_{k k,, l l}^{* *} | | \\ + + {Σ Σ}_{k k = = 11}^{n no} | | {r r}_{k k,, l l} - - {r r}_{k k,, l l}^{* *} | | + + {Σ Σ}_{i i = = 11}^{q q} | | {r r}_{i i,, c c b b} - - {r r}_{i i,, c c b b}^{* *} | | + + {Σ Σ}_{i i = = 11}^{q q} | | {r r}_{i i,, a a u u t t o o} - - {r r}_{i i,, a a u u t t o o}^{* *} | | \\ + + {Σ Σ}_{i i = = 11}^{q q} | | {C C}_{i i} - - {C C}_{i i}^{* *} | | | | 11 - - {r r}_{i i,, c c b b} {r r}_{i i,, c c b b}^{* *} - - {r r}_{i i,, a a u u t t o o} {r r}_{i i,, a a u u t t o o}^{* *} | | \end{matrix} - - - - - - ((22))$

上述表达式中，r_k,m和r_k,m ^*分别表示各个元件主保护的实际状态和期望状态；r_k,s和r_k,s ^*分别表示单个元件近后备保护的实际状态和期望状态；r_k,l和r_k,l ^*分别表示单个元件远后备保护的实际状态和期望状态；C_i和C_i ^*分别表示断路器的实际状态和期望状态；表示连或运算；_ri,cb和_ri,cb ^*分别表示断路器失灵保护的实际状态和期望状态；r_i,auto和r_i,auto ^*分别表示断路器自动重合闸的实际状态和期望状态。In the above expressions, r _{k, m} and r _{k, m} ^* respectively represent the actual state and expected state of the main protection of each element; r _{k, s} and r _{k, s} ^* represent the actual state and expected state of the near-backup protection of a single element respectively State; r _{k, l} and r _{k, l} ^* represent the actual state and expected state of the remote backup protection of a single element respectively; C _i and C _i ^* represent the actual state and expected state of the circuit breaker respectively; _{ri, cb} and _{ri, cb} ^* respectively represent the actual state and expected state of circuit breaker failure protection; r _{i, auto} and r _{i, auto} ^* represent the actual state and expected state of circuit breaker automatic reclosing respectively.

步骤2、求解基于量子遗传算法的改进型中压配电网故障诊断模型，对中压配电网进行故障诊断。Step 2. Solve the improved medium-voltage distribution network fault diagnosis model based on the quantum genetic algorithm, and perform fault diagnosis on the medium-voltage distribution network.

而且，所述步骤2的具体步骤包括：And, the concrete steps of described step 2 include:

(1)采用传统遗传算法对中压配电系统内元件进行编码后，根据量子比特的编码方式进行修正，从而制定适应于量子遗传算法的编码方案用以表示中压配电网故障诊断问题；(1) After using the traditional genetic algorithm to encode the internal components of the medium-voltage distribution system, it is corrected according to the encoding method of the qubit, so as to formulate a coding scheme suitable for the quantum genetic algorithm to represent the fault diagnosis problem of the medium-voltage distribution network;

(2)根据量子遗传算法求解所述改进型中压配电网故障诊断模型并根据计算结果定位中压配电网中的故障元件并判别保护及断路器动作的正确性，进行故障分析。(2) Solve the improved medium-voltage distribution network fault diagnosis model according to the quantum genetic algorithm, locate the faulty components in the medium-voltage distribution network according to the calculation results, and judge the correctness of protection and circuit breaker actions, and perform fault analysis.

而且，所述步骤2的第(1)步的具体编码方法为：And, the concrete coding method of the step (1) of described step 2 is:

假定停电区域整体为个体染色体q，配电系统内元件总数为所述染色体q中的基因个数n；采用量子遗传算法中的量子比特的编码方式，即用一对复数定义一个量子比特位，则个体染色体q采用量子比特编码来解决故障诊断问题的具体形式为：Assume that the blackout area as a whole is an individual chromosome q, and the total number of components in the power distribution system is the number n of genes in the chromosome q; the quantum bit encoding method in the quantum genetic algorithm is adopted, that is, a qubit is defined by a pair of complex numbers, Then the specific form of individual chromosome q using qubit coding to solve the fault diagnosis problem is:

${q q}^{t t} = = [\begin{matrix} {α α}_{11}^{t t} & {α α}_{22}^{t t} & ... ... & {α α}_{n no}^{t t} \\ {β β}_{11}^{t t} & {β β}_{22}^{t t} & ... ... & {β β}_{n no}^{t t} \end{matrix}] - - - - - - ((44))$

上述表达式中，和为复数形式表示量子位对应态的概率幅值；t为染色体的代数。In the above expression, and It is a complex number representing the probability amplitude of the state corresponding to the qubit; t is the algebra of the chromosome.

而且，所述步骤2第(2)步中根据量子遗传算法求解中压配电网故障诊断模型的计算方法，包括如下步骤：Moreover, in the step 2 (2), the calculation method for solving the fault diagnosis model of the medium-voltage distribution network according to the quantum genetic algorithm includes the following steps:

①对中压配电网网架结构进行分析，明确配电网中的元件种类以及数量，确定在故障分析中需要进行分析的元件；① Analyze the grid structure of the medium voltage distribution network, clarify the types and quantities of components in the distribution network, and determine the components that need to be analyzed in the fault analysis;

②根据配电网中开关及保护的动作状态缩小故障诊断范围，确定配电网故障后的停电区域以及需要进行分析的元件；② According to the action state of the switch and protection in the distribution network, the scope of fault diagnosis is narrowed, and the power outage area after the distribution network fault and the components that need to be analyzed are determined;

③根据配电网故障发生后各元件、开关、保护以及断路器的状态，建立元件状态矩阵并根据步骤1中所述的改进型中压配电网故障诊断模型整理目标函数；③ According to the status of each component, switch, protection and circuit breaker after the distribution network fault occurs, the component state matrix is established and the objective function is sorted out according to the improved medium-voltage distribution network fault diagnosis model described in step 1;

④根据步骤2第(1)步的用于表示中压配电网故障诊断问题的适应于量子遗传算法的编码方案设定元件的概率幅值，并对步骤②所确定的元件进行赋值，在数值区间[0,1]之间随机产生一个数，将其与步骤③设定的元件状态进行比较，如果随机数大于或等于概率幅值，则元件的测量结果取1，否则取0；④ Set the probability amplitude of the components according to the coding scheme adapted to the quantum genetic algorithm used to represent the fault diagnosis problem of the medium-voltage distribution network in step 2 (1), and assign values to the components determined in step ②. Randomly generate a number between the value interval [0,1], compare it with the component state set in step ③, if the random number is greater than or equal to the probability amplitude, the measurement result of the component is 1, otherwise it is 0;

⑤将步骤③确定的元件状态值与步骤④确定的元件测量值带入步骤1的改进型中压配电网故障诊断模型的目标函数中进行目标函数评估，确定目标函数初始值；⑤Bring the component state value determined in step ③ and the component measurement value determined in step ④ into the objective function of the improved medium-voltage distribution network fault diagnosis model in step 1 to evaluate the objective function and determine the initial value of the objective function;

⑥设定种群规模、染色体长度、转角步长以及最大迭代次数的量子遗传算法的优化原则并利用量子遗传算法对步骤③目标函数进行优化计算；并将计算结果与步骤⑤中确定的目标函数的初始值进行比较，若该计算结果小于或等于初始值，则保留当前值作为目标函数值；若大于初始值，则更新目标函数值；同时进行算法迭代，直到优化结果满足精度或者达到迭代次数为止；⑥Set the optimization principle of the quantum genetic algorithm of population size, chromosome length, corner step size and maximum iteration times, and use the quantum genetic algorithm to optimize the calculation of the objective function of step ③; and compare the calculation results with the objective function determined in step ⑤ The initial value is compared, if the calculation result is less than or equal to the initial value, the current value is retained as the objective function value; if it is greater than the initial value, the objective function value is updated; at the same time, the algorithm is iterated until the optimization result meets the accuracy or reaches the number of iterations ;

⑦分析计算结果，确定配电网中的故障元件，进行故障研判分析。⑦Analyze the calculation results, determine the fault components in the distribution network, and carry out fault research and judgment analysis.

本发明的优点和积极效果是：Advantage and positive effect of the present invention are:

1、本发明通过分析配电系统内元件种类以及保护配置情况，建立配电网故障诊断模型，模拟故障情景后保护及断路器的动作情况，利用量子遗传算法对故障模型进行求解，从而准确定位故障元件，并判别保护及断路器动作的正确性。实现了配电网故障进行快速、全面诊断的功能。1. The present invention establishes a distribution network fault diagnosis model by analyzing the types of components and protection configurations in the power distribution system, simulates the action of protection and circuit breakers after fault scenarios, and uses quantum genetic algorithms to solve the fault model, thereby accurately locating Faulty components, and judge the correctness of protection and circuit breaker action. The function of fast and comprehensive diagnosis of distribution network faults is realized.

2、本发明能够从配电系统网架结构的角度出发，充分考虑了配电网中的元件信息，保护信息以及断路器信息。对于元件信息，主要分析元件的数量、种类以及保护的配置情况；对于保护信息，主要分析主保护、近后备保护以及远后备保护的配置情况；对于断路器信息，主要分析断路器失灵保护以及自动重合闸的配置情况，进而全面、快速的对配电网进行故障诊断，使配电网能够在故障后迅速排除故障，保证配电网的检修效率。从而为现有配电网故障诊断方法的提供支撑，有利于提升城市配电网规划以及故障诊断水平，促进城市电网建设结构与故障检修手段的合理发展。2. The present invention can fully consider the component information, protection information and circuit breaker information in the distribution network from the perspective of the grid structure of the distribution system. For component information, it mainly analyzes the quantity, type and protection configuration of components; for protection information, it mainly analyzes the configuration of main protection, near backup protection and far backup protection; for circuit breaker information, it mainly analyzes circuit breaker failure protection and automatic protection. According to the configuration of the reclosing switch, the fault diagnosis of the distribution network can be carried out comprehensively and quickly, so that the distribution network can quickly eliminate the fault after the fault, and ensure the maintenance efficiency of the distribution network. In this way, it provides support for the existing distribution network fault diagnosis method, which is conducive to improving the urban distribution network planning and fault diagnosis level, and promotes the reasonable development of urban power grid construction structure and fault maintenance methods.

附图说明Description of drawings

图1是本发明的故障诊断处理流程图；Fig. 1 is a fault diagnosis process flowchart of the present invention;

图2是本发明的实施例中的配电网联络结构图。Fig. 2 is a diagram of the connection structure of the distribution network in the embodiment of the present invention.

具体实施方式detailed description

以下结合附图对本发明实施例作进一步详述：Embodiments of the present invention are described in further detail below in conjunction with the accompanying drawings:

一种基于量子遗传算法的中压配电网故障诊断方法，如图1所示，包括以下步骤：A method for fault diagnosis of medium-voltage distribution network based on quantum genetic algorithm, as shown in Figure 1, includes the following steps:

步骤1、采用元件动作的实际与期望值并融入断路器失灵保护以及断路器自动重合闸的保护状态建立改进型中压配电网故障诊断模型。Step 1. Establish an improved medium voltage distribution network fault diagnosis model by using the actual and expected values of component actions and incorporating the protection status of circuit breaker failure protection and circuit breaker automatic reclosing.

按照一般方法，采用元件动作的实际与期望值可以建立如下中压配电网故障诊断模型，该模型的目标函数为：According to the general method, the following fault diagnosis model of medium voltage distribution network can be established by using the actual and expected values of component actions. The objective function of the model is:

$E E. ((x x)) = = {Σ Σ}_{k k = = 11}^{n no} | | {r r}_{k k,, m m} - - {r r}_{k k,, m m}^{* *} | | + + {Σ Σ}_{k k = = 11}^{n no} | | {r r}_{k k,, s the s} - - {r r}_{k k,, s the s}^{* *} | | + + {Σ Σ}_{k k = = 11}^{n no} | | {r r}_{k k,, l l} - - {r r}_{k k,, l l}^{* *} | | + + {Σ Σ}_{i i = = 11}^{q q} | | {C C}_{i i} - - {C C}_{i i}^{* *} | | - - - - - - ((11))$

上述表达式中，r_k,m和r_k,m ^*分别表示各个元件主保护的实际状态和期望状态；r_k,s和r_k,s ^*分别表示单个元件近后备保护的实际状态和期望状态；r_k,l和r_k,l ^*分别表示单个元件远后备保护的实际状态和期望状态；C_i和C_i ^*分别表示断路器的实际状态和期望状态。In the above expressions, r _{k, m} and r _{k, m} ^* respectively represent the actual state and expected state of the main protection of each element; r _{k, s} and r _{k, s} ^* represent the actual state and expected state of the near-backup protection of a single element respectively State; r _{k, l} and r _{k, l} ^* represent the actual state and expected state of the remote backup protection of a single element respectively; C _i and C _i ^* represent the actual state and expected state of the circuit breaker, respectively.

通过进一步的分析发现，上述基础模型中存在着一定的问题。首先，对于断路器动作状态的期望不能仅仅考虑控制其动作的保护状态，应该由断路器保护范围内的设备及其相关保护的实际状态所共同决定；第二，上述基础模型可能会存在多解问题，在实际的应用中不利于故障诊断的快速、准确进行。Through further analysis, it is found that there are certain problems in the above basic model. First of all, the expectation of the action state of the circuit breaker should not only consider the protection state that controls its action, but should be jointly determined by the actual state of the equipment within the protection range of the circuit breaker and its related protection; second, the above basic model may have multiple solutions It is not conducive to the rapid and accurate fault diagnosis in practical applications.

在本步骤1中，综合考虑上述问题，在该中压配电网故障诊断模型的目标函数的基础上进一步融入了断路器失灵保护以及断路器自动重合闸等保护状态，建立改进型中压配电网故障诊断模型，该模型的目标函数为：In this step 1, considering the above problems comprehensively, on the basis of the objective function of the fault diagnosis model of the medium voltage distribution network, the protection states such as circuit breaker failure protection and automatic reclosing of the circuit breaker are further integrated, and an improved medium voltage distribution network is established. Power grid fault diagnosis model, the objective function of the model is:

上述表达式中，r_k,m和r_k,m ^*分别表示各个元件主保护的实际状态和期望状态；r_k,s和r_k,s ^*分别表示单个元件近后备保护的实际状态和期望状态；r_k,l和r_k,l ^*分别表示单个元件远后备保护的实际状态和期望状态；C_i和C_i ^*分别表示断路器的实际状态和期望状态；表示连或运算；r_i,cb和r_i,cb ^*分别表示断路器失灵保护的实际状态和期望状态；r_i,auto和r_i,auto ^*分别表示断路器自动重合闸的实际状态和期望状态。对于相关元件保护的期望值，当保护动作时r*的值取1，否则为0；对于断路器，当断路器应该跳闸时C*的值取1，否则为0；对于自动重合闸装置，当自动重合闸应该合闸时r*的值取1，否则为0。改进后的故障诊断模型考虑保护元件更加全面，并且考虑了相关断路器的保护，使模型更加完整、有效。In the above expressions, r _{k, m} and r _{k, m} ^* respectively represent the actual state and expected state of the main protection of each element; r _{k, s} and r _{k, s} ^* represent the actual state and expected state of the near-backup protection of a single element respectively State; r _{k, l} and r _{k, l} ^* represent the actual state and expected state of the remote backup protection of a single element respectively; C _i and C _i ^* represent the actual state and expected state of the circuit breaker respectively; Represents continuous OR operation; r _{i, cb} and r _{i, cb} ^* respectively represent the actual state and expected state of circuit breaker failure protection; r _{i, auto} and r _{i, auto} ^* represent the actual state and expected state of circuit breaker automatic reclosing state. For the expected value of the protection of related components, the value of r* is 1 when the protection operates, otherwise it is 0; for the circuit breaker, the value of C* is 1 when the circuit breaker should trip, otherwise it is 0; for the automatic reclosing device, when The value of r* is 1 when the automatic reclosing should be closed, otherwise it is 0. The improved fault diagnosis model considers the protection components more comprehensively, and considers the protection of related circuit breakers, making the model more complete and effective.

所述步骤2的具体步骤包括：The concrete steps of described step 2 include:

(1)采用传统遗传算法对中压配电系统内元件进行编码后，通过引入量子理论中量子比特(qubit)的概念，根据量子比特的编码方式进行修正，从而制定适应于量子遗传算法的编码方案用以表示中压配电网故障诊断问题。(1) After using the traditional genetic algorithm to encode the internal components of the medium voltage power distribution system, by introducing the concept of the quantum bit (qubit) in the quantum theory, and modifying it according to the coding method of the qubit, the coding suitable for the quantum genetic algorithm is formulated The scheme is used to represent the fault diagnosis problem of medium voltage distribution network.

由于配电系统的故障诊断问题重点关注的是受故障影响的停电区域，因此假定停电区域内的总元件个数为n。对于传统遗传算法而言，停电区域整体相当于染色体；而其中的元件相当于基因，采用二进制进行编码，“1”表示该元件发生了故障，“0”表示该元件未发生故障，因此相当于染色体中的基因个数为n。Since the fault diagnosis of power distribution system focuses on the outage area affected by the fault, it is assumed that the total number of components in the outage area is n. For the traditional genetic algorithm, the blackout area as a whole is equivalent to a chromosome; and the elements in it are equivalent to genes, which are coded in binary, "1" indicates that the element has failed, and "0" indicates that the element has not failed, so it is equivalent to The number of genes in a chromosome is n.

而在量子遗传算法中，使用了一种基于量子比特的编码方式，即用一对复数定义一个量子比特位，一个具有m个量子比特位的系统可以描述为：In the quantum genetic algorithm, a qubit-based encoding method is used, that is, a qubit is defined by a pair of complex numbers. A system with m qubits can be described as:

$|\begin{matrix} {α α}_{11} \\ {β β}_{11} \end{matrix}| \begin{matrix} {α α}_{22} \\ {β β}_{22} \end{matrix}| \begin{matrix} ... ... \\ ... ... \end{matrix} |\begin{matrix} {α α}_{m m} \\ {β β}_{m m} \end{matrix}| - - - - - - ((33))$

其中，|α_i|²+|β_i|²＝1(i＝1,2,......,m)。Wherein, |α _i | ² +|β _i | ² =1 (i=1, 2, . . . , m).

量子比特是一个充当信息存储单元的双态量子系统，是定义在一个二维复向量空间中的一个单位向量，该空间由一对特定的标准正交基{|0>,|1>}形成。A qubit is a two-state quantum system that acts as an information storage unit, and is a unit vector defined in a two-dimensional complex vector space, which is formed by a pair of specific orthonormal basis {|0>,|1>} .

因此，基于上述分析，所述步骤2的第(1)步的具体编码方法为：Therefore, based on the above-mentioned analysis, the concrete coding method of the (1) step of said step 2 is:

量子遗传算法最终所得到的适应度即为中压配电网故障诊断模型中目标函数E(x)的值。The final fitness obtained by the quantum genetic algorithm is the value of the objective function E(x) in the fault diagnosis model of the medium-voltage distribution network.

所述步骤2第(2)步中根据量子遗传算法求解中压配电网故障诊断模型的计算方法，包括如下步骤：In the step 2 (2), the calculation method for solving the fault diagnosis model of the medium-voltage distribution network according to the quantum genetic algorithm comprises the following steps:

①对中压配电网网架结构进行分析，其中包括明确配电网中的元件种类以及数量和确定在故障分析中需要进行分析的元件，例如：主变、母线、馈线线路等以及元件的保护配置情况。① Analyze the grid structure of the medium voltage distribution network, including clarifying the types and quantities of components in the distribution network and determining the components that need to be analyzed in the fault analysis, such as: main transformer, busbar, feeder lines, etc. Protection configuration.

②确定配电网故障后停电区域；② Determine the power outage area after the distribution network fault;

在配电网发生故障后，需要根据配电网中开关及保护的动作状态缩小故障诊断范围，确定配电网故障后的停电区域以及需要进行分析的元件，从而使故障分析更具有针对性。After a fault occurs in the distribution network, it is necessary to narrow the scope of fault diagnosis according to the action status of switches and protections in the distribution network, determine the power outage area after the fault in the distribution network and the components that need to be analyzed, so that the fault analysis is more targeted.

③根据配电网故障发生后各元件、开关、保护以及断路器的状态，建立元件状态矩阵并根据步骤1的改进型中压配电网故障诊断模型整理目标函数，为后续的优化算法分析做准备。③ According to the state of each component, switch, protection and circuit breaker after the distribution network fault occurs, the component state matrix is established and the objective function is sorted out according to the improved medium-voltage distribution network fault diagnosis model in step 1, and it is done for the subsequent optimization algorithm analysis Prepare.

⑥设定种群规模、染色体长度、转角步长以及最大迭代次数的量子遗传算法的优化原则并利用量子遗传算法对步骤③目标函数进行优化计算；并将计算结果与步骤⑤中确定的目标函数的初始值进行比较，若该计算结果小于或等于初始值，则保留当前值作为目标函数值；若大于初始值，则更新目标函数值；同时进行算法迭代，直到优化结果满足精度或者达到迭代次数为止；⑥Set the optimization principle of the quantum genetic algorithm of population size, chromosome length, corner step size and maximum iteration times, and use the quantum genetic algorithm to optimize the calculation of the objective function of step ③; and compare the calculation results with the objective function determined in step ⑤ The initial value is compared. If the calculation result is less than or equal to the initial value, the current value is retained as the objective function value; if it is greater than the initial value, the objective function value is updated; at the same time, the algorithm is iterated until the optimization result meets the accuracy or reaches the number of iterations. ;

在本实施例中，以如图2所示的某地区实际配电网结构为例，进行说明：In this embodiment, the actual distribution network structure in a certain region as shown in Figure 2 is taken as an example to illustrate:

1、配电系统网架结构进行分析1. Analyze the grid structure of the power distribution system

(1)明确配电网中的元件种类以及数量和确定在故障分析中需要进行分析的元件。(1) Clarify the type and quantity of components in the distribution network and determine the components that need to be analyzed in the fault analysis.

由图2可知，本实施例中配电系统的联络结构方面以“手拉手”式为主，其内共含有6台主变，6条10kV母线，24条馈线线路，45个断路器(常开断路器15个，常闭断路器30个)，108个保护。It can be seen from Figure 2 that the connection structure of the power distribution system in this embodiment is mainly based on the "hand in hand" type, which contains a total of 6 main transformers, 6 10kV buses, 24 feeder lines, and 45 circuit breakers (usually 15 open circuit breakers, 30 normally closed circuit breakers), and 108 protections.

(2)基于配电网总体结构对网络中的元件及开关进行编号。(2) Number the components and switches in the network based on the overall structure of the distribution network.

将36个元件编号为e₁～e₃₆，其中，母线编号为B₁～B₆，主变编号为T₁～T₆，馈线线路编号为L₁～L₂₄。45个断路器依次编号为CB₁～CB₄₅。108个保护中，主保护、近后备保护和远后备保护各36个，主保护r_1m～r_36m的编号情况为：B_1m～B_6m，T_1m～T_6m，L_1m～L_24m。近后备保护r_1s～r_36s的编号情况为：B_1s～B_6s，T_1s～T_6s，L_1s～L_24s。远后备保护r_1l～r_36l的编号情况为：B_1l～B_6l，T_1l～T_6l，L_1l～L_24l。其中，m、s、l分别表示主保护、近后备保护和远后备保护。因为算例配电网中馈线之间均为“手拉手”联络形式，因此馈线之间互为远后备保护。The 36 components are numbered e ₁ ~ e ₃₆ , among which, the bus bars are numbered B ₁ ~ B ₆ , the main transformers are numbered T ₁ ~ T ₆ , and the feeder lines are numbered L ₁ ~ L ₂₄ . The 45 circuit breakers are sequentially numbered CB ₁ ~ CB ₄₅ . Among the 108 protections, there are 36 main protections, 36 near-backup protections and 36 far-backup protections respectively. The numbers of main protections r _1m ～ r _36m are: B _1m ～B _6m , T _1m ～T _6m , L _1m ～L _24m . The numbers of the near backup protection r _1s ~ r _36s are: B _1s ~ B _6s , T _1s ~ T _6s , L _1s ~ L _24s . The numbers of the remote backup protection r _1l ~ r _36l are: B _1l ~ B _6l , T _1l ~ T _6l , L _1l ~ L _24l . Among them, m, s, l represent main protection, near backup protection and far backup protection respectively. Because the feeders in the distribution network of the example are in the form of "hand in hand" contact, the feeders are mutually remote backup protection.

2、配电系统故障诊断分析2. Fault diagnosis and analysis of power distribution system

模拟算例测试配电系统发生故障，检测到报警信号来自于保护T_3s、B_3l、T_4m、L_9s均有动作，断路器方面CB₁₆、CB₁₇、CB₂₀、CB₁₃、CB₁₄、CB₂₈、CB₂₉、CB₃₀、CB₃₁跳闸。 _The simulation example tests the power distribution system for faults. It is detected that _the alarm signals come from protections T _3s , _B _3l , _T _4m , and _L _9s . CB ₂₈ , CB ₂₉ , CB ₃₀ , CB ₃₁ trip.

因此，根据以上算例配电系统网络拓扑结构得到需要进行故障诊断的故障区域以及故障元件为：B₃、B₄、T₃、T₄、L₈～L₁₅，上述元件对应的状态向量E＝[e₁,e₂,……,e₁₂]；需要进行故障诊断的断路器为：CB₁₆、CB₁₇、CB₁₉～CB₂₆，上述断路器对应的实际状态向量C＝[c₁、c₂、……、c₁₀]＝[1,1,0,1,0,0,0,0,0,0]；需要进行故障诊断的保护为B_3m、B_3s、B_3l、B_4m、B_4s、B_4l、T_3m、T_3s、T_3l、T_4m、T_4s、T_4l、L_8m～L_15m、L_8s～L_15s、L_8l～L_15l，上述保护对应的实际状态向量为：R＝[r₁r₂、……r₃₆]＝[0,0,1,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0]。Therefore, according to the network topology of the power distribution system in the above calculation example, the fault areas and fault components that need to be diagnosed are: B ₃ , B ₄ , T ₃ , T ₄ , L ₈ ~ L ₁₅ , and the state vector E corresponding to the above components ＝[e ₁ ,e ₂ ,...,e ₁₂ ]; The circuit breakers that need fault diagnosis are: CB ₁₆ , CB ₁₇ , CB ₁₉ ～CB ₂₆ , and the actual state vector C corresponding to the above circuit breakers=[c ₁ , c ₂ ,...,c ₁₀ ]=[1,1,0,1,0,0,0,0,0,0]; the protections that need fault diagnosis are B _3m , B _3s , B _3l , B _4m , B _4s , B _4l , T _3m , T _3s , T _3l , T _4m , T _4s , T _4l , L _8m ～L _15m , L _8s ～L _15s , L _8l ～L _15l , the actual state vectors corresponding to the above protections It is: R=[r ₁ r ₂ , ... r ₃₆ ]=[0,0,1,0,0,0,0,1,0,0,0,0,0,0,0,0,0 ,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0].

由此，根据改进型中压配电网故障诊断模型所示的目标函数整理E(x)的表达形式，采用量子遗传算法对目标函数进行求解。优化算法的优化原则设置如下：种群规模取为50，染色体长度取做10，转角步长设定为0.001，最大迭代次数为800次。经过236次迭代后找到E(x)的最小值为9，此时使E(x)最小的元件状态向量为E＝[e₁,e₂,……,e₁₂]＝[0,0,1,1,0,1,0,0,0,0,0,0]，对应的故障元件为：T₃、T₄和L₉。Therefore, according to the objective function shown in the improved medium-voltage distribution network fault diagnosis model, the expression form of E(x) is sorted out, and the objective function is solved by quantum genetic algorithm. The optimization principle of the optimization algorithm is set as follows: the population size is set to 50, the chromosome length is set to 10, the corner step is set to 0.001, and the maximum number of iterations is 800. After 236 iterations, the minimum value of E(x) is found to be 9. At this time, the element state vector that minimizes E(x) is E=[e ₁ ,e ₂ ,...,e ₁₂ ]=[0,0, 1,1,0,1,0,0,0,0,0,0], the corresponding fault components are: T ₃ , T ₄ and L ₉ .

3、计算结果分析3. Calculation result analysis

通过故障后配电网中各元件的动作情况以及故障诊断后的结果，可以分析得到：当主变T₃发生故障后，主保护T_3m拒动，近后备保护T_3s动作，断路器CB₁₆动作；当主变T₄发生故障后，主保护T_3m动作，断路器CB₁₇动作；当馈线线路L₉发生故障后，主保护L_9m拒动，近后备保护L_9s动作，断路器CB₂₀动作。而本实施例的配电系统中常开开关动作是由于需要利用馈线间联络将失电线路上的负荷进行转移，在此不做过多分析。Through the operation of each component in the distribution network after the fault and the results of fault diagnosis, it can be analyzed and obtained: when the main transformer T3 fails, the main protection T _3m refuses to _operate , the near backup protection T _{3s operates} , and the circuit breaker CB ₁₆ operates ; When the main transformer T ₄ fails, the main protection T _3m operates, and the circuit breaker CB ₁₇ operates; when the feeder line L ₉ fails, the main protection L _9m refuses to operate, the near backup protection L _{9s operates} , and the circuit breaker CB ₂₀ operates. In the power distribution system of this embodiment, the normally open switch action is due to the need to use the connection between feeders to transfer the load on the power-off line, so no more analysis will be made here.

综上所述，在本实施例的配电系统中的故障是一个含有元件主保护拒动的多元件故障，利用本发明的基于量子遗传算法的中压配电网故障诊断方法可以准确地进行故障定位并且查找出故障元件。因此本发明对配电网进行故障诊断可以得到准确的唯一解，并且本发明的改进型中压配电网故障诊断模型可以承受断路器、保护拒动等问题，对于配电系统的故障诊断，保证配电系统的良好运行具有很高的参考价值。To sum up, the fault in the power distribution system of this embodiment is a multi-component fault that contains the main protection of the component and refuses to operate. The fault diagnosis method of the medium voltage distribution network based on the quantum genetic algorithm of the present invention can be accurately carried out. Locate the fault and find the faulty component. Therefore, the present invention can obtain an accurate unique solution for the fault diagnosis of the distribution network, and the improved medium-voltage distribution network fault diagnosis model of the present invention can withstand problems such as circuit breakers and protection refusal to operate, and for the fault diagnosis of the power distribution system, It is of high reference value to ensure the good operation of the power distribution system.

需要强调的是，本发明所述的实施例是说明性的，而不是限定性的，因此本发明包括并不限于具体实施方式中所述的实施例，凡是由本领域技术人员根据本发明的技术方案得出的其他实施方式，同样属于本发明保护的范围。It should be emphasized that the embodiments described in the present invention are illustrative rather than restrictive, so the present invention includes but not limited to the embodiments described in the specific implementation manner, and those skilled in the art according to the technology of the present invention Other implementations derived from the scheme also belong to the protection scope of the present invention.

Claims

1. the medium voltage distribution network method for diagnosing faults based on quantum genetic algorithm, is characterized in that: bagDraw together following steps:

Reality and the desired value of step 1, the action of employing element also incorporates breaker fail protection and breakerThe guard mode of automatic reclosing is set up modified medium voltage distribution network fault diagnosis model; The target of this modelFunction is:

\begin{matrix} E (x) = Σ_{k = 1}^{n} | r_{k, m} - r_{k, m}^{*} | | 1 - r_{k, s} r_{k, s}^{*} - \overset{&CirclePlus;}{Σ} r_{k, l} - r_{k, l}^{*} | + Σ_{k = 1}^{n} | r_{k, s} - r_{k, s}^{*} | | 1 - \overset{&CirclePlus;}{Σ} r_{k, l} - r_{k, l}^{*} | \\ + Σ_{k = 1}^{n} | r_{k, l} - r_{k, l}^{*} | + Σ_{i = 1}^{q} | r_{i, c b} - r_{i, c b}^{*} | + Σ_{i = 1}^{q} | r_{i, a u t o} - r_{i, a u t o}^{*} | \\ + Σ_{i = 1}^{q} | C_{i} - C_{i}^{*} | | 1 - r_{i, c b} r_{i, c b}^{*} - r_{i, a u t o} - r_{i, a u t o}^{*} | \end{matrix}

In above-mentioned expression formula, r_k,mAnd r_k,m ^*Represent respectively the virtual condition of each element main protection and expect shapeState; r_k,sAnd r_k,s ^*Represent respectively virtual condition and the expectation state of the nearly back-up protection of discrete component; r_k,lAnd r_k,l ^*Represent respectively virtual condition and the expectation state of discrete component back-up protection far away; C_iAnd C_i ^*Represent respectively to open circuitThe virtual condition of device and expectation state;Represent to connect exclusive disjunction; r_i,cbAnd r_i,cb ^*Represent respectively breaker failure guarantorThe virtual condition of protecting and expectation state; r_i,autoAnd r_i,auto ^*Represent respectively the virtual condition of breaker automatic reclosingAnd expectation state.

Step 2, solve the modified medium voltage distribution network fault diagnosis model based on quantum genetic algorithm, centeringBe press-fitted electrical network and carry out fault diagnosis.

2. a kind of medium voltage distribution network fault diagnosis based on quantum genetic algorithm according to claim 1Method, is characterized in that: the concrete steps of described step 2 comprise:

(1) adopt traditional genetic algorithm to after element is encoded in intermediate distribution system, according to quantum ratioSpecial coded system is revised, thereby formulates the encoding scheme that is adapted to quantum genetic algorithm in order to representMedium voltage distribution network troubleshooting issue;

(2) solve described modified medium voltage distribution network fault diagnosis model basis according to quantum genetic algorithmFault element in the medium voltage distribution network of result of calculation location is also differentiated protection and the correctness of breaker action,Carry out accident analysis.

3. a kind of medium voltage distribution network fault diagnosis based on quantum genetic algorithm according to claim 2Method, is characterized in that: the specific coding method of described step (1) is:

Suppose that power supply interrupted district entirety is for individual chromosome q, in distribution system, component population is described chromosome qIn gene number n; Adopt the coded system of the quantum bit in quantum genetic algorithm, use a pair of plural numberDefine a quantum bit position, individual chromosome q adopts quantum bit coding to solve fault diagnosis to askThe concrete form of topic is:

q^{t} = [\begin{matrix} α_{1}^{t} & α_{2}^{t} & ... & α_{n}^{t} \\ β_{1}^{t} & β_{2}^{t} & ... & β_{n}^{t} \end{matrix}]

In above-mentioned expression formula,WithFor plural form represents the probability amplitude of quantum bit corresponding state; T is for dyingThe algebraically of colour solid.

4. a kind of medium voltage distribution network fault diagnosis based on quantum genetic algorithm according to claim 2Method, is characterized in that: described step solves medium voltage distribution network fault according to quantum genetic algorithm in (2)The computational methods of diagnostic model, comprise the steps:

1. centering is press-fitted Power grid structure analysis, specifies component kind and quantity in power distribution network,Determine and in accident analysis, need the element analyzed;

2. dwindle fault diagnosis scope according to the operating state of switch in power distribution network and protection, determine power distribution networkPower supply interrupted district after fault and the element that need to analyze;

3. after occurring according to distribution network failure, the state of each element, switch, protection and breaker, sets upElement state matrix also arranges object function according to the modified medium voltage distribution network fault diagnosis model of step 1;

4. according to step 2 (1) step for representing the amount of being adapted to of medium voltage distribution network troubleshooting issueThe probability amplitude of the encoding scheme setting member of sub-genetic algorithm, and to step 2. determined element carry outAssignment, the random number that produces between numerical value interval [0,1], the element state that 3. itself and step setCompare, if random number is more than or equal to probability amplitude, the measurement result of element gets 1, otherwise gets0；

5. the element measured value that 4. the element state value of 3. step being determined and step are determined is brought step 1 intoIn the object function of modified medium voltage distribution network fault diagnosis model, carry out objective function evaluates, determine targetFunction initial value;

6. set the quantum genetic of population scale, chromosome length, corner step-length and maximum iteration timeThe optimization principles of algorithm also utilizes quantum genetic algorithm to be optimized calculating to step object function 3.; AndThe initial value of result of calculation and step definite object function in is 5. compared, if this result of calculation is littleIn or equal initial value, retain currency as target function value; If be greater than initial value, upgrade orderOffer of tender numerical value; Carry out algorithm iteration simultaneously, until optimum results meets precision or reaches iterations beOnly;

7. analysis result, determines the fault element in power distribution network, carries out fault and studies and judges analysis.